The value of data infrastructure modernization for the health system [Podcast]

In this podcast, Nordic’s Director of Performance Improvement Kelly Krulisky sat down with Nordic’s Managing Director of Digital Health Kevin Erdal and Chief Medical Officer Craig Joseph, along with Emory Healthcare’s Corporate Director Matt Robuck and Texas Children’s Hospital’s VP of IT John Hamm, to discuss the positive effects a new data strategy can have on a health system’s ability to manage their patient population and business operations and how it might improve the overall patient experience.

 

Show Notes

[00:00] Intros
[01:06] Where the market is headed and why cloud
[02:03] Difference between innovating and modernizing
[03:11] How TCH and Emory are approaching data modernization
[07:45] Criteria that helps organization decide whether data modernization is right for them
[09:00] Where to start and how to prioritize projects
[11:31] What end users want to see with data transformation
[13:05] The innovations and value that data modernization will bring long term
[20:06] Cost and ROI of data modernization – an established platform is key
[27:29] A clinician’s perspective on the financial implications of data modernization
[29:14] The cost of data silos
[31:00] What does this look like for clinical users and health consumers?
[37:24] What is most exciting about data modernization in 2022 and beyond?

vive_podcast_image_1200x628

Transcript

Kelly Krulisky: This podcast is the second in a series on data infrastructure modernization. While the first focused on the technical aspects, this podcast will cover the positive effect a new data strategy can have on a health system's ability to manage their patient population and business operations and how it might improve patient experience. The host today is Kelly Krulisky. I'm joined by panelists Kevin Erdal, Craig Joseph, Matt Robuck and John Hamm. So, to start, I'm going to ask each of my panelists here to introduce themselves, and I will start with you, Kevin.

Kevin Erdal: Thanks, Kelly. So my name's Kevin Erdal. I'm the managing director and practice lead of our digital health team here at Nordic.

Craig Joseph: And I'm Craig Joseph. I'm the chief medical officer and a pediatrician.

Matt Robuck: Matt Robuck. I'm the corporate director of data analytics and integration at Emory Healthcare in Atlanta.

John Hamm: And I'm John Hamm. I'm the vice president of information technology responsible for our clinical and business systems, as well as our data solutions at Texas Children's Hospital.

Kelly: Perfect. So to begin today, we're going to ask Kevin to actually give just a three to five minute overview of where the market is headed in this area and why cloud.

Kevin: Yeah, absolutely. Well, I think the biggest thing is modernizing technology in general, right. Figuring out what makes the most sense for your organization to meet the needs of not only clinicians, but also the patients, more populations at a broader level. So we want to make sure that each organization is breaking things down to ensure that they're moving in the direction that is in line with this institutional strategy. So what we see a lot of in today's world is cloud-based platforms being considered to really help enable some of the innovative activities and really, again, making sure that we're meeting the needs of the clinician, the patient, the consumer. So, that's where, as we're sitting here at ViVE and we're seeing a lot of these net new startups come up and we're seeing all sorts of net new capabilities from a device standpoint, now let's make sure we have the infrastructure to support some of these new capabilities, like I said, to make sure that we're meeting the need of the patient at the end of the day.

Kelly: So in that Kevin, you mentioned both the words "innovating" and "modernizing" a couple of times. So, Craig, would you mind talking to us a little bit about what is the difference between the two?

Craig: In my mind, one of the problems is that we call innovation anything that's new and shiny. And so there's lots of technology that's innovative, but it actually can't really be operationalized. It's actually not very clinically useful. One example I give is something that's way cool. All the doctor has to do to use it is to log out of the electronic health record and then go to a website and then type in their username and password and then find the patient and then find the patient's data, and then really cool things happen. That's innovative, but probably not that useful because physicians are generally not going to do that. And so, really modernizing the way the data are used and putting them within the workflow of the clinical user, I think, is a big difference. And so, I like the latter term a lot more than the more generic innovative or innovation.

Kelly: All right. Perfect. So, Matt and John, your organizations are taking two different approaches to modernization. One on cloud and one on-premises. So I'd like to hear a little bit from both of you on your approach, the two. And John, let's start with you first.

John: Fantastic. So, Texas Children's Hospital has built what we call an enterprise data warehouse over the last 10 years, but most importantly, what we've built is really a governance structure around how we think about data and how we think about data modernization. And that data structure is part of our quality teams and what we call our clinical technology council. It includes physicians and researchers and administrative to support us. One of the challenges that we've run into is the ability for us to really modernize our data platforms, given all of the opportunities that we hear from those governing groups and all those possibilities that feel like we have a hard time getting towards. And so, last year, our executive team and our physician-in-chiefs and leadership team has made a significant investment for us to modernize.

John: One of the challenges when we think about how we're going to modernize is that we need to think differently about how we build the platform of the future. And speed is key. There are other institutions that have built these modern platforms, and they're reaping great rewards in 2022, and we're seeing that, and they've invested over time. We have the opportunity to invest. And so, we've been meeting with some of these leading institutions and we're asking ourselves the question is, how do we move forward quickly? And how do we move forward fast using all of the learnings and try to avoid the mistakes that others have learned along the way?

John: With that, we know that one of the most biggest challenges is around privacy and security, especially in this age. And so, we are leveraging partners who offer on-premise cloud so that we can create that safety feeling, that comfort feeling, as we build these solutions. However, we're launching that in four different areas at the same time because we know that we'll have success with two or three of them, and one of them may not, but we don't have a lot of time to get ramped up quickly.

Kelly: You mentioned learning from other organizations and some of their mistakes. Name one or two of the kind of top mistakes that you've heard from them.

John: Sure, absolutely. So, a couple of the mistakes that we've heard is, one, is IT building out the platform without the proper involvement of our physicians and our researchers and PIs. And so, sometimes what I've heard is IT goes off and builds either these data modernization platforms or builds predictive analytics studies. But then when it's time to say, "Hey, come and use the platform or come use the study," you get the look, the back that says, "Wait a second. We don't understand how you built that model or you've gone off and built this spaceship," I call it the data spaceship, "and we don't understand it. We don't know how to use it."

John: And so, what we've realized is that from the grounds up, we started our approach from that lesson learned, differently, and we actually started a listening tour. And we met with 15 of our researchers, our leadership, and we said, "What is going wrong, and what would you like to see?" And we are building this data platform in step in alignment with that group, and we continue to press expand, but that's probably the number one thing I've heard people get it wrong is trying to build it within IT without that partnership.

Kelly: Super interesting. So, Matt, as mentioned, Emory is taking a different approach. So talk to us a little bit about Emory's approach.

Matt: Well, and I think in talking to John, sometimes I feel like I'm talking to myself because at Emory, we share a lot of the same challenges. And I think for us, we look at it as it's really people and technology. You have to have both. So, when we look at cloud, I mean cloud native, those services that are offered out of the box for us not having to focus so much time on the backend, on the infrastructure side, similar to what you're doing is really important so that we can focus on the actual analytics and driving the value. So we want to take a lot of our resources that are spending a lot of time moving data around and we want to move them more to the front end where they're working closer to the clinician so that they can really drive that value. It's very exciting, and I think...like you, Emory has made a large investment in this in the past year and it's really this inflection point for us moving from just doing the backend pipes, so now we can look more broadly across driving that analytics value at Emory.

Kelly: So I'm sure that this is a decision or a multitude of decisions that organizations across the country are facing. Kevin, what other criteria do you think weigh into helping organizations to make this decision?

Kevin: Yeah. Well, like you just heard from both John and Matt, the biggest thing is starting. So then you get into, how do you actually start? And that is-

Matt: Which is the hardest part.

Kevin: Yeah. Absolutely the hardest part. So, what are those core use cases that the physician, community is going to be ready to, A) share; and then, B) act on? Because we're talking a lot, I think, between the lines a little bit about making some of this data actionable, making an impact and the right data at the right time while we're seeing patients. So that's where I would highly recommend any one of our clients or customers across the industry in today's world is to start with the end in mind, right. I love the listening tour that you brought up, John, to be able to say, "Go listen to your organizations. See what the biggest pain points are." And then back into, what technology do you really need to help solve those specific problems? And there are a number of different tools and there are a number of different platforms to be considered, but don't try to pick the platform first. Try to figure out what the use case you're trying to solve for, is first.

John: That's so true.

Kelly: So, about that, on the listening tour, I'm sure that you got a lot of feedback, a lot of input, and a lot of number one critical projects that need to be addressed. So then, how do you actually get started? How do you determine how to prioritize then with all that input from the organization and then not disrupt or disturb the stakeholders that you spent so much time and care to make sure that you got their input?

John: Yeah, absolutely. We actually started our listening tour not so much around what projects they wanted to do, but what were the challenges. And we ended up with a Venn diagram that kind of listed the challenges around privacy and consent, technology challenges, and really skill sets were the three categories we landed in. And honestly, when we built this slide, I mean, if you want to build a slide that says, "IT sucks," that was the slide. But I think the key was is that we were open and willing to listen to the challenges, and it turned out to be it wasn't IT that sucked. A lot of what landed on the page surface was challenges around just the silos that can be built across the different departments, across the different sections that just create over time.

John: And when we put those all on the slide and really just looked at ourselves in the mirror on that slide, across the board, all the way up to our in-chiefs and EVPs, it was an overwhelming, "Okay, this makes sense. Now we can have the true conversations to how we start to solve these problems." And it was from that point that we led a full-day workshop with all of our research group. There were over 40 people in it. And we heard from them what the future could look like, how to overcome these barriers, it was fantastic. And that really launched us onto what we needed to focus on for the future.

Kelly: Matt, you agreed. Getting...oh, go ahead.

Matt: Yeah. Well, and I was going to say, I think one of the things, and Kevin and I were talking about this earlier, it's connecting those dots. So going just from IT to the clinician, what's the gap for that? It's probably not an ETL developer that's the gap.

Matt: One of the things that we did is we went out and we hired a data literacy resource and we brought in a great product donor to start to begin to serve as that connective tissue so that the needs on the front-end have a place to land. That's not the technical scope first. It's how is the value being driven on the backend? And then we got really lucky because silver lining of both COVID and we made a selection to go to Epic. We're running Cerner today. So we'll be on Epic October 1. So talk about a change to your environment. So this really provided the opportunity for us to completely overhaul our analytics.

John: What a great opportunity.

Matt: Yeah.

Kelly: Craig, I think that you would agree that IT sucks. Just kidding. But no, you don't necessarily care about the mechanics of data transformation. What do end users want to see out of this?

Craig: Before I answer that, let me tell you I've served multiple times as a chief medical information officer, and I'm willing to change my hat from IT to clinician as need be. So, if there are physicians with pitchforks, I do say, "Oh, yeah. It's IT's fault." And then when IT's upset with the doctors, I'm like, "Yeah, it's definitely the doctor's fault."

Matt: It's a good place to be.

Craig: Yeah. It saved my life a couple of times. In response to your question, I do think that clinicians, in general, don't care about any of this, right. So, on-prem, ask your average doctor or nurse or respiratory therapist whether the data should be housed on-prem in a server or not, they will just look at you and say they don't know what you're talking about. And so, we want it to work. That's all.

Craig: And so, how it...99.99. How many nines do you want there? I just want zeros. I want one and with a bunch of zeros at the end. So it just needs to work and it needs to be responsive and it needs to have the information that I need. And how the infrastructure works behind the scenes, it's to me kind of like plumbing. We don't see what's behind the wall in our house and we don't really care about it until something breaks. And then we say, "Ooh, we probably should have put some more money into that infrastructure when they were telling us that there were some problems." So, certainly we support that, but how it happens and kind of the expertise, way beyond us.

Kelly: All right. Matt and John, what innovations will your data modernization efforts enable long term? And what value will they bring to your organization? So, Matt, let's start with you this time.

Matt: Yeah. And I'll go with this shiny object here, and that's predictive models. Both with Epic's deployment infrastructure, and then what we're bringing in with the cloud, it's going to provide that opportunity. And we're an academic medical center, so we have the research side. So this is the first time that we'll be able to take what the research side's doing and roll that into clinical care, and then have that feedback come back to the research side to complete the circle. So I think that's one of the things from an innovation standpoint that we're really excited about. We have five predictive models that we plan on rolling out at Epic go-live. And then we have a list of a lot more that we already have our eye on. So, I'm very interested-

Kelly: Are you able to tell us what they are?

Matt: Deterioration index, sepsis, things like that, that we're looking at to roll out first. And some of those are models that maybe Epic has built, and some of those are custom models because, again, the great thing about being an academic medical center is that you have a bunch of data scientists on staff. So this is going to be that first time that they can really roll out at the enterprise level, the stuff that they're working on. So, very excited about that.

Kelly: That is very exciting. Do you already have that kind of governance and feedback loop set up from that research side to the clinical side?

Matt: No. And that's a challenge. And we talk about people and technology, the people part and the governance piece. So we're currently building out that structure right now and it's really great to see the involvement the whole organization is getting behind this effort because they see the value and they see the future here. So, I'm very excited, but that's something that we're having to build while we're building the technology.

Kelly: John, tell us a little bit about TCH.

John: Sure, absolutely. So, when we came out of our workshop, we had other meetings and we stood up a project structure. And some of that structure included our leadership team, included that research work group that we depend on heavily now for feedback. And we broke our lanes up into four lanes around the innovation. How are we going to change? The first and foremost that we realize we have to come face-to-face was with privacy and governance because at Texas Children's, our partnership was with Baylor College of Medicine and they lead our research arm as well as our physicians are sponsored at Texas Children's.

John: So the first thing we are doing is we've built a joint management committee between Baylor and Texas Children's and then we're including our legal privacy and compliance teams to be part of the conversation so that we could start to connect there.

John: The second place that we're really focused on is skill sets. So we've worked with UT Health and University of Houston, and we're really thinking about building a curriculum on how we onboard our PIs and researchers so they understand how to get to data. And then we're also starting to talk about potential courses and programs around that could support a fellowship program or others. So it's really around people skill set.

John: The third lane is the process. So, today, if you want data from IT, you submit a ticket. The problem is that when you talk to our researchers, a lot of times when they submit an IRB, which is required for a human based study or a grant, they really need help upfront. And so, we are re-imagining, what if it's not a ticket, but what if it's an Apple Store-like experience where we have a team of people that can be scheduled and they can help just have that upfront conversation around the type of data that's available in Epic, the right tools to use? Kind of help point them and guide them. And we have gotten support to hire a team of 12 people to build that Apple Store-like experience and support our researchers.

John: Finally, is the tech. And when you think about the tech, it's really around launching. And what we have learned is I could spend a lot of time on one shiny object. I love the shiny object. We are launching it in four spaces. One is around Epic Cosmos and Epic's Nebula predictive analytics study. The second area that we're launching is with what we're calling our HPE Ezmeral service, which is the on-premise model. The idea there is to give the researchers tools to be able to explore data on our network. The third is a company called Melax who does natural language processing of your unstructured data. So we want to build one of the largest natural language processing to pull out human oncology phenotypes and do genetic studies. So we're in partnership with our neurological research institute for a business use case there.

John: And then finally, we're working and talking to Google about launching a focus on image study with our pathology team, and that's early in the subjects. We know if I launch those four areas and we do it in partnership with the governance, that one of them may not completely make it, but I'd rather do that than launch just one area and find out in three years that we've completely missed the boat on the opportunities that have the speed. And so, we're really excited about, that's kind of our four prong approach that we've put teams in place to focus on the launch of this.

Kelly: That sounds like a lot. Now, to kind of make it a little bit more manageable, in each of those areas that you mentioned, is there some imaging oncology? Is there a particular like a single use case, for example, in each of those areas that you're working to prove out?

John: Absolutely. And it is challenging and which is why your plug is we have brought Nordic and a team to come in and help us through this. And they've been fantastic. Yes. So, that's absolutely. So, within our research work group and research leadership council, we've identified one to two use cases. And one of the key premises of the use case is that if we're successful with the use case, it's accomplishable, but it also is scalable. And so, we've done that in each of the areas and we have teams that are focused, and we've actually making very good progress on taking the steps necessary to execute.

Kelly: Matt, did you want to redirect?

Matt: Well, I was going to say, Kelly, you're asking the questions, but I'd like, John, to ask you a question. Back to the Apple Store experience and those 12 people, can you talk a little bit more about that and what you're building there and the skill set of those people?

John: Yeah. So, first and foremost, is leadership. So we're hiring a director of research, data analytics, and machine learning. That person will be responsible for working closely with a partner who runs our research administration within Texas Children's, to support them. We have five data scientists that will focus on understanding the data models and partnership with the data team I already have in place, the more technical data team. And then we're hiring five bio-statisticians and their sole job is going to be to provide the support that's needed equitably across the departments. And then there's one administrator who's going to help us just build the front-end websites and process flows that help the researchers actually be able to connect. And we're building the Apple Store experience on Microsoft platform.

Matt: Wow. That's great.

John: Yeah. Yep.

Kelly: So, again, all sounds super easy. Kevin, will you talk us through a little bit about the costs of this data modernization and how should organizations think about that? And if you could add a little bit about too, then putting that in terms of ROI.

Kevin: Yeah, absolutely. So, I would just start off by saying the cost of putting something off really isn't an option anymore, right. I think historically we were talking a little bit about what technology investments do I need to make or where do I start? Right. And I think from both Matt and John, what you've heard is, there's a plethora of technology that we're going to have to bring to the table. So, from a sheer ROI standpoint, that's always very difficult and John, Matt, I would be curious on your feedback from a dollars and cents standpoint. What I typically go back to is the VOI, right. What's the value on investment that you're now able to bring to, again, the patient and the provider community? We know that there is a big resource consideration broadly across healthcare and not just within IT, but also in the clinical practices as well.

Kevin: So if we can leverage some of this technology and we can start to bring technology together, to your earlier point, John, how can we now start to really support the patient community, the populations, and being able to ensure that our clinicians are working on what they wanted to be working on, right? Caring for patients. And if there are administrative tasks, we can reduce or hopefully eliminate even better. If there are some clinical research studies that we can execute on and start to make a difference on how we care for patients, those are the value props that I think I would start to look at. And also why we went back to where I started is with the use cases, right. Each organization, to be able to say, “Hey, where can I make the biggest difference?”

Kevin: So, from a dollars and cents standpoint, I know always very, very difficult to be able to talk to that. We are talking about, in some of the cases, monetizing some of the development activities within healthcare, not just data as an asset, but also some of the predictive models being utilized. Using that as a revenue stream, I think there's great opportunity there, but there are some considerations along the way.

Kelly: You two are the ones who do have to manage the dollars and cents. So, how is that going for your organization?

Matt: Well, and I was just going to say real quick to that point, you can't really monetize if you don't have the platform to build on. There's no even option to monetize.

Matt: So, for us, that's huge. And we're really looking at that as the first step to even be able to get to the table where we can say, "Now we want to build and then release externally." We can't do that today without that platform. So, for us, that's a huge piece of it.

John: Yeah, I absolutely agree. I mean, I think one of the things that, just thinking about what you just said is, during COVID, I don't know how you feel, but I feel like COVID kind of forced our clinicians and leadership and corporations to address problems and challenges using technology very rapidly. And what I'm feeling is that the leadership teams really became digital partners overnight, and they really learned both the challenges and the opportunities that technology can bring in data. I think also during COVID, you saw other health organizations, other, call them competitors, all of a sudden rise with these products. And so, I think the competitive landscape has become really real in technologies at the forefront. And so, coming out of COVID, I think that's in everybody's brain and they're saying, "Okay. This is real and it's possible. And are we moving fast enough?" So I love the analogy of you've got to build the platform to come to the table.

Matt: Well, and I would say too that what we saw is that the questions being asked of the technical side of the house became increasingly more complex. Before COVID, it was a lot of just putting the pieces together, building the blocks. We knew what to do. When COVID happened, you would get these requests. And we were like, "Hey, we don't have the technology to meet that." So, I think that's one of the things that's also shifted for us as an organization to say, "Okay. We need to be ready for this in the future with having the platforms in place."

John: Absolutely. Absolutely. It's interesting because I was talking to one of our lead researchers and she asked the question. She said, "Are we moving too fast on building the platforms?" And I shared a story. I said, back in 2014, I joined an oil and gas company and our CIO was very progressive and we built out data lakes and research networks for our geophysicist. And it was interesting because it took two years, but there was a couple of use cases. And then in 2016, oil prices dropped. No one would've, I mean, oil's $100 a barrel and it just dropped to about $50. All of a sudden, every oil and gas company started looking for data solutions, but they had to go build the road. They had to build the platform.

Matt: Right, right.

John: The place I was at, we were ready. And all of a sudden, that platform was there, it was available, and we led in decisions because of those platforms being available. And she said, "Ah." She goes, "Yeah, that's exactly what we need is to build this platform while we're building up our investing on our research side." Directly on the dollars and cents, organizations measure the dollar and cents through indirect funding as part of grant submissions. And then the value cycle of that is you use that money to reinvest in better, because all of our goal is to improve care for, for Texas Children's, it’s women and children. And so, we know that investing on better quality IRBs leads to higher quality grant submissions. It's good for branding. It's good for the outcomes. It's good for quality. It's good for patient care. It has good across the board. We believe that that investment will cycle itself back and then we'll re-use it again for better care, better outcomes, better improvement. So that's where our focus is at.

Kevin: Yeah. One of the things you're hitting on there, John, too, is reinvesting in that same strategy, right. You have the technology underneath it, so when the next study comes up, the next grant is approved and you have some additional dollars. You're not changing your strategy. And that's where we have to get away from that siloed mindset or if there's another device that comes into an organization, well, that's fantastic. Let's enable that device, whether it's for remote patient monitoring or whatever the scenario may be. Great. We're investing into that same exact strategy. We might be adding technology, but that doesn't mean we're necessarily changing course. And I think that's really important after listening to both of you here.

Matt: Yeah, I agree. And from an Emory perspective, we’re looking at some ROI on the operations side as well through predictive modeling. We have one predictive model that we’re planning to roll out in radiology around incidental findings, where we have about $6 million. Not only do we have patients who need to come back in, but we have about $6 million in revenue that we’re leaving on the table because of some backend manual processes. So, we see some gains on the operational side from an ROI perspective as well. I’m very excited about that and very excited to actually measure those as we grow into this because I think predictive models, sometimes the front-end clinical piece, but sometimes the backend side of it too from an efficiency perspective and from a revenue perspective, there’s a lot of gains to be had there as well.

Kelly: Not surprising. I think speaking as more of an operational user and audience, it is constantly shocking to me how little we do that is through data-driven decisions or kind of data-informed workflows. So, not at all surprising to understand that there might be that ROI there from an operational perspective as well.

Matt: We're in an organization of over 24,000 people. There's going to be some inefficiencies in that. And if you can use technology specifically around predictive models to help kind of smooth those edges, there's gain to be had there as well.

John: If you don't mind real quickly, I'm just interested Dr. Joseph, when you hear organizations talk about this from a financial perspective, with your clinical background, your focus on patient safety, quality, how do you react when you hear the conversation being about finances when it comes to research or building these type of platforms? Just curious.

Craig: Yeah. Well, no margin, no mission. And so, it's pretty straightforward. We all want to do the best that we can for every patient that we're seeing. And I think that's true from the CEO on down to your hospitalist and to the primary care docs that are referring patients in. We also realize that money is what it takes to get the hospital off the ground and to keep it running and to do all of these fancy things. And so, I don't think any of us are turned off by it. We realize it and we all want a new wing for whatever specialty that we are, and we all want the latest equipment and all of that, but certainly money is the, as long as it's being well spent, it makes sense. And I think one of the things that you've done successfully is with that listening tour and painting that picture of, "Hey, this is what we'll be able...this is what we think we're hearing from you. And these are the kinds of things that we'll be able to do. And this is the way that we'll be able to do it." "Ah, now I understand."

Craig: And I'm going to continue to go back to my This Old House analogy and talk about people who are repairing or building houses, not wanting to put the money into the infrastructure, not wanting to put the money into things that are behind the wall. But when you explain to someone that, "Hey, if we don't this, we know that 10% of the time within a number of years, this bad thing is going to happen. How do you feel now?" Oftentimes those answers are changed. And so, certainly, it's all about communication and being transparent.

Kelly: Craig, kind of to now take away the analogy. In your many capacities, you've served obviously, as you said, informatics IT, as a physician, as an executive. What have you seen data silos actually cost organizations?

Craig: Oh, well, we're having lots of problems with burnout of our clinicians and some of it's because we're certainly asking them to do more in less time. A lot of it though is that there's so much information that's out there, but it's not readily available. And one example is data that is outside of our organization, but we're aware of it. I'll give you an example of mammograms. Oftentimes the payer, the insurer, knows that a mammogram happened, didn't happen at our organization. They know because they paid for it, right. So they have that information and then we have access to their claims information. So, we as the hospital or healthcare system, have access to that. We don't know the results. We don't know what it said, but we know that they did have it. And so, those are the things that I see physicians and nurses and others spending time, "Okay. When was your last mammogram? Okay. You think it was five years ago? Okay. We're going to need to spend money that doesn't need to be spent necessarily.” If that information could be surfaced, it would be so much easier. And I think that's what we're really looking for.

Craig: We're out of the area of physicians not having access to the data. Now we're at the problem with physicians having access to too much data, right. They can now see almost everything and they're responsible for seeing almost everything and no human could possibly do that. And so, what we're really looking for to our technology partners to be doing is to be finding that information that we think is important and surfacing it to us at the right time. And that sounds super easy and it's really quite complicated.

Kelly: So I think that that's a nice segue into kind of the clinical applications. So, clinical applications on modernized data infrastructure. What does this look like for clinical users and then health consumers? Let’s start with Matt.

Matt: Well, yeah. And I'll just piggyback off of what I was talking about earlier with the radiology incidental findings app. So being able to deploy an application like that at scale, at enterprise is something that we can't do without the cloud. So being able to put that across the Emory enterprise, so both have the patient impacts on the front end when they get scheduled to come in for a visit that they should have come in for. And then on the backend, for the physicians to have that data at their fingertips. I think having the cloud is what really enables that deployment for us of those types of applications.

Kelly: Absolutely. What would you have done?

John: So we have a number of use cases. And just to give a little bit of credit for one that I'm really excited about, is there was a study that's being fostered out of Children's Hospital of Philadelphia. And in our neurological research institute, there's a Dr. Liu. Dr. Liu is leading a study in partnership with Children's of Philadelphia around identifying epilepsy cases earlier in the process. And what we started off within his research study was that he needed us to extract patient records related to patients that have had epilepsy to try to identify, more quickly identify the type of epilepsy, and I may be saying that wrong, Dr. Joseph. The type of epilepsy earlier in the process.

John: My understanding is that when a child comes in as having seizures or onset seizures, it can actually take up to two years to really find the right medication chemistry that is best for that child. As determining that today, a lot of it is based off the physician experience and it's based off of the trials and the onset and learning more about that patient, completing studies, trying on different medicines.

John: Long story short is, one of the research use cases we're going to try with Melax and again, Nordic, is helping us through this process, is actually what they found is that the unstructured data in the EMR has more accurate data and more timely data than depending just on the ICD codes. Thus what Melax has proven, born out of UT Health, is that they can identify human oncology phenotypes associated with epilepsy, and they can score them. Then if you pair that with the genetic mutation test data, they believe, and what they've shown, is they can more accurately share with the physician and predict the type of epilepsy that this patient might be suitable, and most importantly, what type of medicines might be more successful or treatment program.

John: And so, what our first use case and parallel is building out this Melax program, but then also working closely with the physician through the study. The reason I love this one is because if we're successful, then we don't want to just build this natural language processing capability for one use case, this is scalable into our cancer oncology center. This is transferable to our heart center. There's so many cases. And so, we've already established. And back to your point, Dr. Joseph, the vision is not to complete one study. The vision is be successful here, learn from it, and build one of the largest natural language processing human oncology phenotypes, and share that knowledge with our researchers, have the program in place to train them on how to do that, so they can take that back to their sections and repeat these type of studies because we really believe that could be a game changer.

Matt: That's fantastic.

John: Isn’t that cool? Yeah, that one gives me chills.

John: Very cool, yeah.

 

Matt: That’s a cool one.

Kelly: Craig, you are a physician and a pediatrician. So, anything that you would want to comment further on what John just shared or anything that you just want to comment further on what this looks like for clinical end users?

Craig: The main message that I would love to have IT folks learn, and I'm going to reference John's Apple Store. Oftentimes people come to me when I have that CMIO hat on and they say, "This is what I need. I need a new order set that does this. I need a new documentation flow sheet. I need a new documentation template that looks like this," and they'll actually draw it out for me. But often when I say, "What is the problem that you're trying to solve?" It's a different thing altogether. And they don't need those things. They know what those tools are, and they're trying to make the problem solvable by that tool and it just doesn't work. And there are other tools that they're not aware of.

Craig: And so, you don't generally walk into an Apple Genius Bar and say, "I think I need a software update 42.16 because that's going to help me a lot." You say, "It doesn't work." "What's the problem?" "It used to work, and I'd like you to get it back to the way it was working before." And so, really kind of coming from a clinical or operational standpoint, coming to our IT partners and saying, "Hey, this is the problem that I'm having, which is what both of you have done. And I'm not going to tell you how to solve it because you're the specialist in IT. I'm the specialist in operations or clinical care. Please meet me halfway and suggest how we should solve this." And you're generally going to come up with a better answer that way.

Kelly: I think that really resonates. I think that we all understand what it's like to kind of get lost in the land of tickets. You mentioned earlier that’s such a common process these days and Craig and Kevin and I are used to kind of sitting in between operations and IT. And the first thing that we do is not to comb through all of the tickets. So I think that that resonates across the board, but Kevin, did you want to add something there?

Kevin: Yeah. I think just breaking down some of what Craig was hitting on, right. From an IT perspective, we need to listen and then we need to make sure that technology isn't the reason we're doing the right thing for the clinician, for the patient. And that's really what it comes down to. If I boil everything down at the end of the day, listen first. Make sure technology isn't the problem. How do we do that? We want to have modern platforms, make sure we have connectivity, give that 360 view of the patient for the physician to really ultimately help the patients.

Matt: Solving clinical problems, not technology problems.

Kevin: That's exactly right. Very well said.

Kelly: So we have talked a lot today about things that are to come and these are things that will take probably years to really evolve. And we'll go back to the house analogy since that's one we're sticking with today. A lot of what's happening today is really around that building infrastructure. But, that being said, here in this year, 2022, what are you most excited about what is likely to happen for you and your organization or for the healthcare industry in 2022 related to this data modernization?

Matt: So I think it's obvious probably at this point for me. October 1 is our Epic go-live. Having five predictive models into production this year, enterprise at scale, is just very exciting because this is a long time coming for our organization. Like I said, we've got a lot of the talent on the university side. This is kind of the realization of, and kind of that inflection point going forward. So, for us, that's going to be huge.

Kelly: Should we distribute your phone number or Instagram handle here so that people can text you on the first and see how it's going?

Matt: It won't be going well for me, but it'll be going well for Emory.

Kelly: Yes.

John: That's right.

Kelly: Perfect. What about you, John?

John: I think the most exciting this year that I’d like to see is the alignment across our leadership team on focusing on this. I mean, we really see some movement. We have a new pediatrician-in-chief that came from Boston. She's really working closely, she's aligned to building a better strategy. Our CEO and CIIO, Meyer Davis, is focused on this strategy. We have support across the board from our physicians. Just seeing that collaboration and some of those conversational walls break down is absolutely the most exciting.

John: From a pure delivery perspective, we've signed ourselves up for quite a bit. So we expect to have our pilot use cases; at least one pilot use case across those four platforms completed by the end of our fiscal year, which is in October. And so, I want to see that. And then we'd like to do one at least use case where we try out this Apple Store. And then, of course, we're building a team. We're hiring 12 people in a market that's really hard to find people. So, it can't just be about salary. It's got to be about people that want to be part of this exciting mission. And so, being able to build out that team. I think those are all really realistic activities. If I see that happen, it will just be a great year.

Kelly: This is exciting, Matt. So the end of your fiscal year is 9/30. Of course, your goal, however, is 10/1. So, I feel like we have to reconvene on October 2.

Kelly: But I think we're all eager to hear the feedback and eager to hear the feedback on the Apple Store as well. So that will be exciting. So, Kevin and Craig, you're going to bring a different point of view. More from an industry perspective, what are you most excited about for 2022?

Kevin: Yeah, for me, I think it's the adoption of a lot of what we're talking about right now, right. So modern platforms and some of the technology we're talking about isn't drastically new this year in 2022. What we're really getting into is the ability to start to adopt some of the processes and start to adopt some of the technology we've all been working on for a period of time. Instead of the house analogy, I like the highway analogy where I feel like we're on this four-lane highway, we're going down the road, and we're so close to be able to get to our favorite restaurant, which is just off to our right, but there's no off ramp. Well, now we're starting to build that off ramp. Now we're starting to be able to get there. And now that's where we can start to make the difference. Bringing the right data, the right people at the right time, so to speak. Make that last mile, if you will, achievable. That's what I'm most excited about.

Kelly: Craig, I meant to have Kevin go last to force him to come up with an original idea, but you got stuck with it instead.

Craig: That's all right. I am most excited about ambient listening. That does not sound very exciting. I think we are getting closer and closer, and I'm not sure we're going to hit it in 2022, but we're getting close to be able to listen to a doctor interacting with a patient in a clinic room, doing multiple things. Having the computer actually get very close with the progress note that is pretty decent. It may not be perfect, but it really does capture the high points and the important things.

Craig: And secondly, being able to queue up orders and send messages to folks. So, if I were to say, "Hey, I think little Johnny's got a bad ear infection and it's not going away. We're going to start him on an antibiotic. Let's start him on amoxicillin at the usual dose." To have that order queued up to refer to a patient. So that's the science fiction that doctors have been dreaming about for decades now. And I think we're getting close. So that's the technology that I'm pretty excited about.

Kelly: And that is not to totally segue us into a completely different topic, but you obviously mentioned provider burnout earlier. That is a huge problem right now. And that what you just mentioned will go a long way to trying to address a lot of the problems resulting from that. So, very good call. Before we close, any final comments or questions that anyone would like to say?

John: I'll just say, I appreciate the partnership with Nordic and thank you for hosting us on this podcast, and I look forward to the future. It's fun to have a vision of the future and then actually be part of an organization that has the leadership to execute it. And that's where I'm at, where I feel like I'm at Texas Children's Hospital.

Matt: Yeah. And I would say same here. Earlier in the year, we bit off a lot with Epic, with this platform, and I think Nordic's partnership in all of this has made that possible. I'm less scared about those dates than I was maybe six months ago. I'm still scared, but I have confidence that we have the right team in place, both on the Emory side and the Nordic side.

Kelly: We need to rest up and do some meditation now.

Matt: Oh, yeah.

Kelly: It's not going to happen later in the year. All right. Well, I want to thank each of our panelists for their participation today, as well as our listeners for joining us. If you do have additional questions, want to learn more, or want to discuss further, please visit us at nordicglobal.com.

Topics: featured, digital health, Interoperability

Module heading text

Get the highest quality chemistry and microbiology testing services aligned closely with current good manufacturing practices (CGMP) for all types of products across all phases of development.

Subscribe to receive blog updates