Episode #34

Podcast with Seth Hain, VP of R&D and Sean Bina, VP of Access and Patient Engagement, Epic

“We essentially see ourselves as stewards helping clients manage their data”

paddy Hosted by Paddy Padmanabhan

In this episode, Seth Hain, Vice President of R&D and Sean Bina, Vice President of Access and Patient Engagement at Epic discuss the next wave of opportunities for Epic and how the company has evolved by focusing on patient experience and advanced analytics.

At Epic, the focus has always been on providing patients with access and tools to view and have control over their data. Epic works with over 300 health systems today to help them manage their data. Over 160 million consumers have or are using its MyChart patient portal which has been around for nearly two decades. The company uses advanced analytics such as AI/ machine learning and monitors how they are performing on different populations before embedding it into workflows, be it clinical-facing or patient-facing.

Epic has lately started focusing on providing transparency around healthcare costs and has been working on creating accurate estimates for patients so that they have price transparency at the point of care.

Welcome to The Big Unlock podcast where we discuss data analytics and emerging technologies in healthcare. Here are some of the most innovative thinkers in healthcare information technology talk about the digital transformation of healthcare and how they are driving change in their organizations.

Paddy Padmanabhan: Hello again, everyone and welcome back to my podcast. This is Paddy and it is my great privilege and honor today to introduce my special guests. We have two of them today, Sean Bina, who is a Vice President for Patient Experience with Epic and Seth Hain, who is VP of R&D for Epic. Seth and Sean, thank you so much for joining us and welcome to the show.

Sean Bina: Thanks for having us.

Seth Hain: Thank you.

Paddy Padmanabhan: You are most welcome. So why don’t we get started? Maybe you can give us a little bit about your background and your current roles at Epic for the benefit of our audience.

Sean Bina: Yeah, so I’ll start. So, this is Sean and I started at Epic 23 years ago doing implementations of our systems. And then over the years, I’ve worked with a variety of our different products and now focus my time on the patient experience. And that’s really my goal, to help patients get connected into their health and wellness in a way that they’ve never been able to before.

Seth Hain: My name is Seth and I started a little after Sean. I’ve been here for about 15 years and when I came to Epic, my focus was on architecture and the kind of systems infrastructure behind the scenes. And then I combined that with my prior experiences around mathematics and focused on and continued to today the research and development around analytics and machine learning and in particular embedding that type of intelligence into workflows, be it clinical facing or patient-facing, for example. And the tools that are used throughout the system for analytics and machine learning.

Paddy Padmanabhan: And between the two of you, I would argue that you are looking after the top two focus areas for health systems today – patient experience and advanced analytics. So, I’m looking forward to this conversation. I want to start with this, now we have near-total penetration of electronic health record systems in the country today. Of course, it’s been a great run for Epic as leaders in the market. So, what are you seeing as the next wave of growth opportunities for Epic?

Sean Bina: So, we see a few different areas where we’re continuing to do a lot of work. One is we just continue to work with our customers to expand the adoption of Epic and then to add new additional modules. One of the things that we’ve learned over time is that every specialist needs a system that’s really designed specifically for them, whether they are core things that all physicians and clinicians use around ordering and reviewing results and doing some basic documentation. But then there’s a lot of subspecialty support that’s needed for doing things like managing images, for registry support and for doing the kind of specific documentation within a given subject area. So, we continue to go deeper and deeper into those areas to create a great experience for physicians. The one thing I would add is when we look at the industry in general, we do see our customers are really focused on three core areas. One being patient experience, two being analytics, but then three is really bringing the joy back to practice for physicians.

Seth Hain: I would also add that the space around healthcare continues to change and evolve. And I think there’s a lot of opportunities as we start to look at and see more collaboration between, say, payers and providers in the space. There is an opportunity there to help facilitate faster exchange that benefit both the provider and their workflow, but also the patient and the care they’re receiving in a timely manner. And that expands as you start to think about a broader definition of health. Thinking about things like dental, long term care, even where people get care, be it a telehealth encounter through, you know, they pay from the app directly on their phone, or they’re in a retail clinic and they just need to swing by for a flu shot and making sure that’s a continuous experience from a health perspective.

Sean Bina: The other thing that we continue to work on is we continue to move internationally. So, you know, outside the U.S., we work with an increasing number of countries on our software and there’s kind of a wonderful cross-pollination that we get as a result of that. So, for example, we’ve been working a lot with Finland and they’re very focused on social determinants of care, really, really focused on reducing the need to get into the hospital or readmissions into the hospital. And so, a lot of the work that we’re doing with them around their social care system then ends up benefiting our U.S. customers.

Paddy Padmanabhan: Right. Right. And we’ll unpack some of these, especially around the emerging tech stuff in the context of digital transformation, which is what we mostly focus on as a part of this podcast. So, let me ask you a simple question. What is your understanding of the term digital used by everyone? And almost everyone has a definition for it. And how do you define digital? What is your understanding of the term and how is it impacting all your choices or investments in health systems, specifically from your point of view with your clients?

Sean Bina: Yeah. If you don’t mind, we take a quick trip back down memory lane. You know, when I started back around 20 plus years ago, the world was very different in terms of what was available from a digital perspective. You know, there weren’t integrated ambulatory inpatients solutions. You couldn’t do an end to end revenue cycle system that covered all your hospital and clinics. And so, people were using best of breed systems and trying to cobble systems together to manage all of that. And until around 2003, there wasn’t even such a thing as a patient portal. So, one of the things that when I think about what it means to be digital today, it’s obviously changed over the years. But I think it’s important to remember how far we’ve come that we now do have fully integrated systems that cover all the kind of food, warmth, and shelter that is needed by healthcare organizations. So that now we’re getting to the point where we can do a little bit more of the poetry.

Seth Hain: I think that depending on who you ask or the definition of digital transformation, you tend to hear a pretty different perspective. Some people immediately come at it from the patient experience perspective and the possibility of having access to health care through your phone at any point in time. Others and I tend to take a foundational view of systems perspective on some of these topics. You know, start to also bring in things like cloud computing behind the scenes and the running of machine learning algorithms on data that is flowing into the system from a combination of devices. Then being able to be back into wherever the provider might be so that they can be sure that in the ICU, be that walking down the hallway so that they have a better-informed picture of what patients they might want to spend some time with at that exact moment in time. So, I think that gets to more of the transformation point in building on what Sean was saying.

Sean Bina: Yes one other key element of digital is the interoperability piece. So, if you go back to the year 2005, 2006, 2007, interoperability was a fax machine. We’re now passing five million records a day around the country and it’s starting to be around the world for patients. And so we’re really starting to see where physicians have gone from thinking of records from the standpoint that I have my record at this site versus there are other records at other sites to wanting to have combined digital views of all of a patient’s information pulled seamlessly together.

Paddy Padmanabhan: Yeah, I think we certainly have seen a significant amount of progress on interoperability. But I just wanted to make one observation. You know, when you look at the past decade or so, well, the single biggest thing that happened in terms of digital transformation was really the digitization of medical records. I remember working with paper records with my physician 10 years ago, and it so happened that he was fighting tooth and nail about going digital. But if you really look at it, I don’t think any of the sectors has seen the kind of transformation that healthcare, in particular, has seen, just by virtue of digitization of medical records. So now it looks like we’re, you know, phase one of the mission has been accomplished. Now they have the strong foundation of digital records. And so, everyone seems to be talking about what do we do next with it? Advanced analytics, building better experiences, looking at data from multiple sources and so on. So, in that context, the whole competitive landscape is also changing. So I want to probe a little bit on how you see Epic evolving in the context of this emerging landscape of technology players and the evolving needs for health systems, as they compete with a whole different marketplace with a lot of nontraditional competitors in all kinds of other things going on. Do you want to comment on that? How are you evolving and what are the changes you’re seeing your clients go through and how are you evolving in step with that?

Seth Hain: You’re getting back to that kind of world of transforming from a paper chart to a kind of maybe a desktop PC where somebody would go to get information, I think is an interesting analogy to kind of transformation we’re seeing right now where in many cases it’s not about the chart, right? It’s about a continuous health experience that folks are receiving. Be that a patient or a provider. So, we see somebody like Rush down in Chicago who builds automated workflow to understand as patients come into the E.D., their likelihood to leave based on machine learning algorithm using inputs from a variety of different sources, not just the medical record, and then use that to help drive workflows where they can walk around, touch base with the patient, let them know where they’re at from in my perspective, to see a nurse or a clinician. And they saw a drop of about 50 percent in folks leaving without having been seen from the emergency department. So, you start to see a different type of transformational workflow emerge that isn’t based around a single machine but is more driven by a backend kind of ubiquitousness of data accessibility from a cloud perspective. And then differing devices be those iPads being used for rounding or watches used to alert physicians in the ICU of patients that might be at risk of deterioration from based on a machine learning model. So, it really starts to transform how clinical practice is being given some of the financial aspects as well, are also are being looked at.

Sean Bina: Yeah, I would just add from the patient perspective, I think they don’t think in the same way as they used to that I have a record at a particular healthcare organization. Increasingly, they’re thinking about their health and wellness and how that includes what’s going on their Fitbit, and on their Peloton and on the medications that they’re taking. And so, they’re looking at a much broader ecosystem of inputs. And I think the expectation is growing that what healthcare organizations are going to be able to do is take all of those inputs, pull them all together, and then provide recommendations based on a much broader set of data than that’s ever been data available in the past.

Paddy Padmanabhan: Yeah. And my firm did some research back in the summer of 2019. On the current state of digital transformation in healthcare, and what we found was that over half of the health systems that we polled in the study were looking at the electronic health record platforms systems as a starting point for the digital transformation. One of the reasons was that integration aspects in pulling data adding it all from within the workflow of an EHR system is easier to do. And when you talk about digital transformation, people are talking about integrating data from multiple sources. But it’s still a lot of integration work that is involved here. I want to switch at this point to talk about the data itself. You know, your obvious strength for Epic is in the data that you have access to, all the patient records that are being processed through your system, across all these health systems across the country. Now, that is a huge advantage to Epic as you try to build out your models and as you try to build out your experiences and just get a better understanding of your patient populations. Can you share a couple of examples of how you’re actually using the access to the data to improve experiences as well as outcomes for patients? Do you want to talk about one or two examples? Maybe you mention a couple of clients where you’re doing some work in this regard?

Seth Hain: Yes, sure. So, I can address a couple of those points. One of the things to be clear here about is that organizations work with us and we essentially see ourselves as kind of stewards helping them manage the data that they have on-site. And we work with them to kind of build out workflows that have the opportunity to be fully informed by the data in their system and the context around the patient and the provider. We often think of this internally as a concept we call relevance, where we want to make sure that full picture is brought to bear. I mean, some of the easiest examples to think about in this context are around the acute space where we rapidly see folks deploying machine learning models around things like sepsis, deterioration, and fall risk being three of the most common ones. We see folks start with often implementing them as a bundle. And at this point, we have over 300 organizations either running directly in their system or in the midst of implementing machine learning models in those types of contexts. And it’s exciting to see the impact that it has. And, you know, it ranges from something like a 17.7% decrease at the North Oak, which is a community hospital down in Louisiana in mortality reduction for sepsis patients to, you know, also a decrease in alerts that providers are saying you use machine learning models to better identify patients. It also helps save folks time. So, UC Health, who spoke to Amy about this, saw a 19 percent reduction in the number of alerts they were seeing in these types of contexts as well. So, it both benefits from a provider time saving perspective while improving care.

Sean Bina: We also do models on the operational side of the house. So, doing things like identifying the patients that are most likely to “no show.” So, in the past, you obviously could run kind of massive report and you could do a lot of analysis and trying to find this information. But now we can’t just have the system waiting. What are the most important variables and identify which patients are the most likely no shows and then do things automatically based on that information? So, whether it’s doing a reminder phone call or texting the patient or whether it’s overbooking the patient because they’re unlikely to show up at a particular day in time. We can automate some of those processes. For me, that’s part of the excitement as you mentioned, as we’ve got completed the underlying digital transformation; we can now do these things. One of my favorite examples is what we call a FastPass at Epic, what it does is it automatically monitors the waitlist, identify the patients that are at the highest risk and need to get in the soonest, and then will automatically text or email them when new appointments become available. John Hopkins, they saw about a twenty-seven-day improvement from when patients were scheduled with a specialist until when they got in based on using this FastPass. So at one time improves the convenience and access for patients, but then at the same time, it also helps the healthcare organization because you’re filling times that would have otherwise gone unfilled or where you would have had to have a lot of staff managing the situation.

Seth Hain: And to your point, Paddy, I think, about platform, that equally important to data in regards to machine learning and these types of scenarios is the workflow in understanding what data is present and ready to be used at the point in time that somebody can make an intervention that will really matter and how to get that information into the people’s hands. They can do something with it. And so, as we build out more machine learning models here in the data science team at Epic, that’s actually where we start. It’s not with the data that’s available or those sorts of things, but it’s about the impact that we want to have in the workflow and how we see that fitting in and then work back towards the true kind of machine learning training processes and the stuff that the data scientists really do day to today.

Paddy Padmanabhan: Yeah, there was this one question that I have on this. When I talk to CIOs and digital transformation leaders, one thing that I hear often is that it’s very important to identify the right kind of use cases if you will, or AI II and machine learning applications. And often I’ve heard that the bigger opportunities today, maybe in more in non-clinical use cases, administrative functions, and revenue cycle management as an example or even for the patient experience related applications. Is that the sense you get in based on all of the work that you’re doing that we’re further ahead? Or maybe there’s a bigger opportunity in the short term with non-clinical versus clinical use cases? What would you have to say on that?

Seth Hain: I would hesitate to say that that is exactly what we see. I think it’s different depending on the area. Certainly, in the operational areas, we see real opportunities for automation, and we see folks using machine learning embedded into the workflow to save folks time and energy in regards to moving through those operational workflows. On the acute side, we see a variety of impactful outcomes like the ones I just referenced, be it around deterioration, be it around sepsis, be it around palliative care. There is a lot of opportunity that folks see there and documented outcomes such as on the North Oaks, one that I shared a moment ago. There is also real opportunity in the population health space. I think it is harder there to truly measure the outcomes when you’re looking out two to three years in regard to the impact that folks have and directly tying it to the clinical management and care management that takes place. That’s not to say that it doesn’t help. You know, we have documented evidence and dug in as we build machine learning models around, say, sorting outreach for diabetic patients just based on an A1 C value compared to a predictive model of their two-year risk of Type 2 diabetes complications. We can really see a difference in the math when we dig into that. It’s harder and takes longer to produce those studies, though, about outcomes longer term.

Paddy Padmanabhan: Yeah, yeah. But let’s switch to the patient experience. Patient access, patient experience, these are hot topics for health systems today, high focus areas. And there’s also a teaming ecosystem of a digital health startup funded by billions and billions in venture capital money that are addressing specifically the patient experience and patient access aspects of the healthcare value chain. Now, you know, when I look at health systems, I ask myself – what would we be looking beyond an electronic health record system for? As I mentioned earlier that half of the health systems that we polled in our study were already using electronic health records systems for most of their digital functionalities. But there is also a growing trend of using startups. So, I guess this is a question for Sean. Where do you see Epic fitting in this overall milieu of digital startups that are coming up with maybe new ways of defining experiences and new solutions? Where do you see Epic fitting in this overall context?

Sean Bina: Yes. So, first of all, a little bit of context. MyChart has been around for a long time now. It goes all the way back to 2003 when we first went live with it. And kind of the patient side has always been a focus for us at Epic in terms of providing patients with access and tools to be able to see and view their records. That’s always been a core thing that we want to make as seamless as possible. And we now have almost I think we’re over one hundred and sixty million MyChart accounts. So, we’re closing into the point where about one in two people in the country having and are using a MyChart account today. And we’re starting to see much greater adoption than in the past. So, I think in the past, you know, we would see our customers have around a quarter to a third of their patients be active MyChart users. But the trend is way up. And so, for example, one of the most interesting things I’ve heard recently is that at M.D. Anderson, if a patient is seen three times, then there’s a 90 percent chance that they’re using MyChart for M.D. Anderson. Now, of course, those are patients that are sick, and they have a whole set of issues. But what it shows is that patients really will adopt the technology when it comes to using it, when they do have health issues. And so, I think one of the questions for us is not will patients use MyChart if they’re sick and they have chronic diseases and they’re in for surgeries and all of that. We know that for patients that are connected to health systems, that they will become active adopters of MyChart. But for people that are generally well or have particular health concerns are just trying to manage their health issues, but are not constantly going in to see the doctor, how do we reach out and get connected to those patients? And so, our focus is really kind of turning to help patients do a lot more self-management and do a lot more wellness within the system than we’ve ever done in the past. Some of that is providing people with targeted education based on the information that’s flowing into Epic. So, when you talk about this whole ecosystem of startups, a lot of startups feed information in which we can then consume and take advantage of in Epic. So, whether it’s your blood pressure monitor, your heart rate monitor, whether it’s your Fitbit, whether it’s your Apple Watch, all of those things then become data feeders that then get, consumed through MyChart up into the EHR and then we can provide monitoring and management of that data based on configuration within Epic.

Paddy Padmanabhan: Yeah. And you know, one of my recent guests on the podcast mention that right now we have a fairly low threshold for option of digital tools by patients. If people start using the tool, that itself is a significant change. And it’s really heartening to hear that you’ve got 160 million patients who are now beginning to actively use MyChart in some way. I am one of them, by the way, and I can’t remember the last time I actually called into my physician’s office for scheduling an appointment or just for non-emergency type questions. I do that all through MyChart today.

Sean Bina: Yes. I’m the same way and I use MyChart all the time for managing my daughter’s care and then managing my care. And so, you know, I love to access digital tools and I would much prefer to always do something online than have to make a telephone call. And I think many, many patients are in that same boat. There is a cultural change that still needs to happen at many health care organizations to give patients more control. So, one of the things that I’m continually advocating is that we don’t need to wait to give patients their test results until after a physician has reviewed them. We should be providing open notes to patients as much as possible. And so, we have those capabilities within MyChart today. And so, it’s just a matter of transforming the healthcare system. And some of this will almost certainly be mandated by the government in the next year or two is that really to provide that full context for patients when they go into their shared medical record.

Paddy Padmanabhan: Yeah and that would be a huge leap, actually, especially the comment you made about the notes and all that. So, let’s talk about emerging tech stuff again. You know, we talk mostly about digital transformation and now we are on the cusp of some big breakthroughs with some of the emerging technologies that can potentially play a big role in the way health care is delivered in the future. So, we just touch on a few of them. And let’s start with this one, cloud computing. What are your thoughts on the role of cloud going forward in digital health?

Seth Hain: I think there’s a number of things that cloud computing provides, but at the end of the day, I think it is really about faster delivery of technology to folks to be able to put it into practice. So, a couple of years ago we released our cloud-based machine learning platform, which is essentially provided as a service and allows organizations to embed directly into their workflows, machine learning algorithms that run in real-time on the latest data in the chart. And when we built that out, we build it out in a manner that used. Forgive me, I’ll dive into a little bit of techno-jargon here, but used containers, which is a kind of new approach for deploying software out on the cloud and is agnostic, so that runs on Microsoft Azure today can run on other platforms as well in the future. And that enables organizations to also both getting access to new things we’re developing here in Verona, but also to embed their own software more efficiently. So, we’ve seen organizations like Ochsner who now have deployed nine different machine learning models directly onto that platform and embedded them back into their workflows. So, they see this as a tool to allow them to more rapidly evolve both their clinical and financial operational workflows. And they share those types of results and approaches our UGM conference, XGM, and in other forums so that folks can learn how to do that and move more quickly with it. So, I really see cloud as an approach for faster delivery and that then enables that type of faster execution on new clinical programs and the like.

Paddy Padmanabhan: Yeah, I had a quick follow up question on that. So you mentioned Microsoft Azure, so they are the big tech firms that have their own plans for the healthcare market, in some ways they may well be competing with you, and in other ways, you could be partnering with them. So how do you approach this today at Epic? Where are you partnering with? What do you think you’re going to be competing and what do you see as it relates specifically to the big tech? I’m talking Microsoft, Google, Amazon in particular.

Sean Bina: Its really customer driven. So, you know, its what customers are coming to us and asking us for and then us doing an evaluation on our side in terms of what models are going to work the best, who’s the best groups to partner with?

Paddy Padmanabhan: Right. OK, good. So, let’s move on to the next one on my list – voice recognition. Boy, I have to tell you, I’m pretty excited about what I see in terms of its potential. I just saw a news item that said that we now have the ability to identify biomarkers based on voice. And I thought that the future of health is here. But anyway, I don’t want to get too far ahead of myself. What do you guys think?

Sean Bina: Oh, we’re super excited about voice too. So, you know, people have been using voice obviously for years in terms of using systems like drag and then model to capture notes. And we have a lot of physicians that are highly efficient doing that. But we certainly want to kind of add a few additional layers onto them. And so, the first thing that we did was we started creating a voice assistant. We now have a voice assistant that runs on our mobile platform where I can say, hey Epic, and then have it answer certain sorts of questions for me. And then we are doing the work to move that into hyperspace. So on the workstation, you’ll be able to have a microphone where it will work in the ambient fashion and you’ll be able to use voice commands to drive workflow, to find out information about a patient, and to really work hands-free. I think a lot of our focus is, you know, where does this technology make sense? Whether it’s, you know, in the room for an inpatient where the patient is the driver, whether it’s in the OR, where people are scrubbed in or whether it’s in the clinic where the physician is focused on the patient instead of being focused on the workstation. And how can they quickly get the information that’s most relevant and then get things cued up in a simple and easy fashion? And we feel like the voice assistant is going to be a great way to do that. And then a little bit farther, but not that much farther down the path is the conversational capture with diarization and natural language understanding to basically be able to start to construct a note out of the natural conversation that is happening during a visit. We’re already seeing kind of experimental groups are doing this in areas like orthopedics where you have pretty structured common visits that are happening over and over again and then using machine learning to eventually get to the point where instead of having a human being as your virtual scribe, the system is really the virtual scribe creating that note.

Paddy Padmanabhan: Yeah, ambient clinical computing environment I think that’s kind of becoming a term in vogue today. I saw one of my earlier guests on his podcast mention that in 10 years’ time we’re going to completely keyboard-less and we’re going to have a voice-enabled or ambient computing environment where you don’t need a keyboard anymore. That’s where physicians are looking forward to that because that is going to significantly reduce their burden. But, how close or how far away are we? Is voice recognition mature enough today? What are the error rates within control, what’s your quick comment on that?

Sean Bina: Kind of we’re learning a lot right now is what I would say. So, we know that voice recognition works when I have a microphone in my hand and that they are at 95, 96, 97 percent accuracy using the new cloud computing platforms. So, for example, when Dragon moved to the cloud, the accuracy increased, and it can handle more accents and different styles of speaking better than ever before. So, we know that the accuracy is really good in kind of that clinical scenario. We’re kind of in the first layer of watching how the voice commands are being used and that hit rates and success for that. And we’ll learn a lot as more and more customers go-live. And then, you know, we’re kind of doing close monitoring on these first areas where people are piloting ambient voice assistance in specialty areas like orthopedics. So my sense is that in areas where you have a fairly structured dialog, you’ll have fairly fast adoption over the next year to two years in areas where a kind of a more classic internal medicine visit, where a patient might have nine different problems and you have a 45-minute visit where you’re covering all different types of things with the patient and doing a lot of education. But that’ll take a little bit longer.

Paddy Padmanabhan: Yeah. Yeah. Well, I’m still trying to get my car to listen to me and play the exact song that I want to hear. And so, I am, I guess, a little further away on this. So, the last one on my list for emerging tech and we’ve covered a fair bit of this with Seth as far as, you know, artificial intelligence. But I do want to touch on one thing as it relates to that topic. I hear a lot from people who are practitioners in the field and the customers as well that there are some concerns, they have about black-box algorithms, algorithmic bias, even some ethical considerations around the use of AI in certain contexts. So, Seth, do you have any thoughts on how we address these and where we are in really gaining or enabling customers and users to gain more confidence in these tools?

Seth Hain: Two things really, I think come to mind. The first one is really understanding how they’re being put into practice, where machine learning is really embedded into this system to kind of augment the information available to a user. I have a quote here from a user group meeting presentations from Denver Health, where they were talking about implementing a deterioration index model and a nurse shared with the folks, then put it in the plot practice that the deterioration index doesn’t change the way I nurse my patients, but it gets me into the right rooms faster. Right. So, understanding how those are embedded into the workflows I mentioned previously, I think can certainly go a long way to addressing it. The second piece that comes to mind is really about the process that an organization goes through as they implement I and put models into practice. First, understanding how it is performed on similar populations. What went into building the models? We published briefs on every model we create that organizations can review prior to putting it into their system. The second is the ability to run that model silently and understand how it performs at the organization in the context that is likely to be used in prior to putting it into practice. The third is obviously putting it into the workflow and making sure that users understand the context in which is being included and having it embedded directly there in an explainable fashion.

Sean Bina: And actually, this is true not only for AI/ machine learning, this is true for all decision support. You can’t turn on the decision support alert without first running it silently seeing when it’s going to be triggered. How often it is going to be triggered and whether it’s being triggered in the right circumstances? And then you have to measure over time how a decision support alert is being used for it to be effective.

Seth Hain: And I think that last point is key here. It is often not talked about in the context of machine learning, where after you have machine learning models live and in practice, it’s important to continue to monitor them and understand how they’re performing on different populations and then taking steps to adjust that where appropriate. It might mean adjusting the model. It might mean adjusting the workflow. But understanding its performance on a variety of different individuals, and in a variety of different circumstances over time is key and we provide that type of monitoring capability directly within the platform for augmentation.

Paddy Padmanabhan: Yeah, I think that makes a lot of sense. We’re coming up to the close for our podcast here. Is there anything that you’d like to share with our listeners about any new product features or new functionalities that you planning to launch this quarter or the next quarter?

Sean Bina: One thing that has been available for a while that I would just quickly highlight because we didn’t talk about it. We didn’t talk very much about the financial side of the house. But transparency around how much healthcare costs is absolutely essential. We have to make it so that as patients are coming in for their visits, they have a good sense of what this visit or procedure is going to cost them, what’s going to be out-of-pocket, what’s going to be covered by their insurance, and provide them with an understanding of the mechanics of that. In many cases, a patient might even be charged differently going to the same healthcare organization, depending on whether they’re going to the clinic or the hospital for the same disease. So, we have to make sure that patients know about that as they’re making these decisions about when and where to go in and who they’re going to be seen by. And so, we’ve been really focused on being able to create estimates for patients that are highly accurate based on historical data and can be provided at the point of care by the doctor. So, the doctor can say this is what an upper GI is going to cost. He or she can say what the medications are going to cost and whether there are less expensive alternatives. And then also providing that same information as a patient is going in to schedule their visits and procedures. So, to me, that’s a really big deal and it’s something that could really transform a patient’s experience by not being surprised by what the costs of things are in the end.

Paddy Padmanabhan: Yeah. And I agree completely with you. It is a big deal. I think the price transparency, cost transparency is something that is unfinished business as far as healthcare is concerned. Well, Seth and Sean, it’s been such a pleasure speaking with both of you. I greatly appreciate your joining us on this podcast. And I look forward to catching up with you again soon. Thank you again.

Sean Bina: Thank you.

Seth Hain: Take care.

We hope you enjoyed this podcast subscribe to our podcast series at www.thebigunlock.com and write to us at info@thebigunlock.com

Rubik Cubes make a perfect gift for any occasion. Learn more about this entertaining puzzle visiting ruwix.com.

About our guests

Seth Hain, Vice President of R&D at Epic, focuses on integrating analytics and machine learning into healthcare. This includes the development of business intelligence tools, data warehousing software, and a platform for embedding machine learning across Epic applications. During his 13 years at Epic, Seth has also led the Systems and Performance group, with an emphasis in database performance and architecture.

A native of Seward, Nebraska, he received a BS in Mathematics from the University of Nebraska and an MS in Mathematics from the University of Wisconsin. Seth currently resides in Madison, Wisconsin with his wife and two children.

Sean Bina is the Vice President of Access and Patient Engagement at Epic. His focus is on improving health and wellness by helping people to become more connected, knowledgeable, and in control of their care. He currently divides his time between strategic application planning and product management. During his 23 years at Epic, Sean has worked as an account manager, team leader, implementer, an RFP writer, and as a salesperson.

Sean graduated from Beloit College with a degree in Philosophy and Literary Studies. He lives with his wife, daughter, and dog in Madison, Wisconsin.

About the host

Paddy Padmanabhan is a widely published and quoted thought leader on digital transformation in healthcare. He is the author of  The Big Unlock: Harnessing Data and Growing Digital Health Businesses in a Value-Based Care Era, and the CEO of Damo Consulting Inc, a digital transformation and growth advisory firm based in Chicago.

Connect

Share on facebook
Share on twitter
Share on linkedin
Share on email

Subscribe to our podcast