Episode 11 : Reimagining Healthcare with User-Centered Design and Agile Methods

SFTS_Thomas Maddox.mp3: Audio automatically transcribed by Sonix

SFTS_Thomas Maddox.mp3: this mp3 audio file was automatically transcribed by Sonix with the best speech-to-text algorithms. This transcript may contain errors.

Intro/Outro:
Welcome to Smart From the Start, presented by Care.ai, the smart care facility platform company and leader in AI and ambient intelligence for healthcare. Join Steve Lieber, former CEO of HIMS, as he interviews the brightest minds in the health providers space on truly transformative technologies that are modernizing healthcare.

Steve Lieber:
Hello and welcome to Smart From the Start. I'm your host, Steve Lieber, and it's my pleasure to bring to you a series of conversations with some of the sharpest minds in health information technology. We'll discuss the smart directions healthcare companies and providers are pursuing to create smart care teams. Today, I'm joined by Dr. Thomas Maddox. Dr. Maddox is the Vice President, Digital Products and Innovation at BJC Healthcare at Washington University School of Medicine in Saint Louis. In this role, he provides strategic oversight and direction of both the BJC-WSM Healthcare Innovation Lab and the BJC Digital Team. He is also a Professor of Medicine and Cardiology at the Washington University School of Medicine. Welcome, Dr. Maddox.

Thomas Maddox:
Thank you. Appreciate the invitation.

Steve Lieber:
Excellent. Looking forward to talking with you. It's always a treat to have someone who has at least a connection to the medical preparation, education, training, and such that really feeds the future of healthcare. And a lot of our conversations around technology obviously is focused on the future. So let's talk a little bit about your read today in terms of what you're seeing, in terms of the people coming out of medical preparation and what is it that we need to be focused on today in preparing them for entering clinical practice that might be different than ten years ago, 15 years ago?

Thomas Maddox:
Sure. I appreciate the question. I think to start with, it's I really enjoy the work that I do on an academic medical center. And I'll say that obviously most of health care is provided through community-based health systems, which is wonderful. But I think we in the academic medical center world have both the advantage and the responsibility, really, of providing some of those newer models of care and doing the research that really provides evidence that those new models really do help really make a meaningful improvement. Obviously, the patient outcomes, but also to the clinicians experience, because we know that the pressures and the intensity of a medical career are only increasing. And we also know that technology has promise to potentially help with that. But in order to understand the right mix of technology and practice, we're going to need innovation labs like the one we have here to experiment with different care models and really bring some of that research rigor that schools of medicine like Wash U have to truly understand the benefit of those models. The final thing is, I think any current graduate of a medical school is by definition a digital native. They've never grown up, they're in their 20s, so they've never grown up without having the internet, without often having at least some form of digital device. And so the idea that they would come to us conducting every other aspect of their life in a heavily digitally enabled way and not have that be a core feature of how they practice medicine. I think it would just be strange and really inefficient and ineffective. So I think efforts around both providing the digital-enabled care models and having them participate in both design and execution is just a requirement in today's day.

Steve Lieber:
You hit on what my first follow-up question was going to be about digital natives in terms of the entering physician workforce, and you just validated that they are coming in with that background doesn't even seem sufficient enough. It's just the way of life, the way they work. It's the way they do everything, as we all have and think about. Over my career, it was predicted that this was going to happen. There will be a time when the incoming physicians, nurses, other clinicians just come to expect it to be there versus it being introduced to them, as we have with previous generations of clinicians.

Thomas Maddox:
That's right. That's exactly right. You know, when I feel around in the hospital, I treat inpatients with cardiac issues. And what I have developed is a habit that I think is really useful, is I'll often bring our technology teams and their background is not medical. It's their designers, their product managers, their technologists, but they are a little younger. They tend to be in their 30s. And I'll bring them with me on clinical rounds, and at the teaching hospital where I round, I usually have house staff with me, medical residents, and fellows who are also in their late 20s, early 30s. So they are contemporaries and what I have them do is join us on rounds so that the technologists get a sense of what is it like to be on the front of patient care and how do we think about it? And how do we organize our work to be able to do that well? And when they look at it through their technologist's eyes, I'm like, why do you do it that way? Or that seems redundant, or what's a fax machine? Other things like that where we're like, no, no, this is part of the reality. And concurrently what they'll do is start talking to their contemporaries on the medical side who say, oh yeah, we have to do these tasks to take care of this patient with our disease. But they already know that they're not doing everything they could digitally. So they'll often ask the technologists, they're like, it would be really great if I had this information that flashed on my phone as I was walking into the patient's room, and it already knew the information I needed to have an effective conversation or design the management plan for this patient. And often what I find is that the crosstalk that our young docs and our young technologists have is just so rich, and they come away with so, so many new ideas of varying degrees of practicality, but it just serves as fodder, and I think speaks to where innovation is going to come from, and that's from our next generation of technologists.

Steve Lieber:
This cross-talk you just mentioned, sounds like it could be part of the concept behind your Innovation Lab, because I think that crosses the boundary of the facility and the School of Medicine. Talk a little bit about that Lab, some of the things you're working on, and what has the promise there, because you mentioned it's an important component that especially in an academic medical center focused on the research and ensuring that practices are, in fact, validated and clinically sound. So talk a little bit about that lab for us.

Thomas Maddox:
Sure. Yeah. I think when the Lab was started, our vision was we know that a lot of digital technologies were maturing and making their way into healthcare, but we weren't yet sure about where it could be best applied and to what end, to optimally support a better experience and better outcomes for the patients and the care teams that are taking care of them. With that overall focus area, we then categorized the way people get care in three different areas. One is people need access to care, or we providing the digital tools that tell them where you can come for care, how you can engage with us, can you have a virtual visit, those sorts of things. The second is and actually the delivery of care itself. And are we appropriately digitizing and using all the information that people bring to us when they're in a clinic or they're in a hospital, so that we know as much as we can about them and their needs, so that we're designing care plans with them that are going to give them their best shot at optimal health. And then the final thing was recognizing that the average person spends far less than 1% of their waking hours in the health care system. They're not sitting in a clinic every day, they're not hopefully not sitting in our hospital every day. They're living their life. But we know that what they do in their life directly impacts how much they're going to need us as a healthcare system. Are they smoking, are they exercising? Are they working in a hazardous environment? Are they in a good social or not a good social situation? So all of these various factors that we know in aggregate contribute to health and will affect the healthcare that they need, can often be visualized and in some cases advised by digital tools. We can start monitoring what people are doing with wearables and activity levels, with air sensor monitors for pollution levels, with sensors in the home to understand, some of the activities that they do that may impact their health. So starting to think about what tools would be useful and practical to help inform that part of their lives, to inform healthcare that's been part of the work. Very practically, we've done a fair amount in our so-called digital front door to improve the information we give our patients and ease their path in getting access to us. We've done a fair amount in both predictive analytics and digital point of care tools to be able to anticipate the kind of care that somebody might need, and then to surface the information for their care team to allow them to effectively treat that patient.

Thomas Maddox:
And then finally, we've done a fair amount of work in so-called remote patient monitoring and using wearables and in-home sensors, primarily with people with chronic disease like diabetes or heart failure to keep tabs on their health. And if they're showing early signs of deterioration, raising the flag for help with their care teams. That's been a lot of it. We've also, like everybody, started to probe into the newer capabilities that are coming out of digital health. Generative AI has a lot of exciting potential, but we're in super early innings there. We also know that some of these digital tools are not only going to be useful for helping patients get good care, but starting to unburden our physician and other healthcare providers that right now are screaming for help in terms of the information overload and just the absolute lack of bandwidth, they have to get their work done in a timely way, and that's really impacting their wellness epidemic levels right now, burnout. And there's probably a role done well for technology to help offload some of that burden. So those are the various areas that we've currently got in our portfolio.

Steve Lieber:
Excellent. Let's touch on that last one. Because the impact on the workforce of life today, the clinical life, the home life, all of this stuff, it leads to some serious challenges. And we do we see burnout and a number of issues in the past. We thought about the need to get out of paper into the digital world. So we were focused really on data collection. That's the way I visualize what the EMR was all about was collecting data. Then we started analyzing it, and next generation was around digital analytics and such. Now you've introduced the generative, the ability to apply actually a new generation tool onto that data that's been collected, some degree of analysis. But now we move in even beyond predictive into generative. One of the I hear some sort of pros and cons sometimes in the conversations. Some clinicians are looking at this as a threat. Others, as you just articulated, it has opportunity to relieve the burden we can use ambient monitoring to. And then AI on top of that to predict which ones we really need to focus on when because of signs and signals. Is it a pro and con? Is there a downside and an upside here in terms of where this next-generation technology is going as it relates to the workforce? Is there something we need to do to prepare them and help them realize this is if we do it right, this is going to be a help?

Thomas Maddox:
Oh, absolutely. I think there's a lot of potential. I think if it's done without deep thought, it will be. I don't know that it'll be harmful, but it'll certainly be less effective than it could be. I would like to tell you categorize sort of our evolution of digital. I'll call it capabilities and information over time. The way I thought about it is we have data collection or digitization, and the drive to give everybody over the last 15 years has been a big part of that. But I would argue the next steps are data curation, data insight, and then data generation. And I think all three are still fairly underdeveloped. And I think if we do this well, all three will address some of the variables that drive physician burnout from a data curation point of view. We have so much information, particularly if you're older and you've had some medical care over the course of your life. If I'm seeing you as a new patient in the hospital or the clinic, I still in a digital way, but I still have to manually comb through your charts. Now there are digital charts. They're no longer the paper charts when I was training, but we still have to comb through them, and they're often an instruction just not easily retrievable. And I think there's a lot of opportunity for us to continue to both organize our structured data in a good way and develop our natural language processing tools that can go and effectively mine the unstructured data, the clinical notes, and the conversations that people have, and synthesize all that into a concise and sort of form factor that I, as a physician, need to truly say, okay, Steve, it's coming to me. And over the past 60 years, here are the various things that have happened to him that have allowed, that have impacted whatever medical condition or medical issue he might be coming to me with, or that predicts that he may run into this down the road. So let's maybe introduce some preventative measures to mitigate or forestall anything happening. We are getting there with predictive analytics, but I think there's a lot more work to do. I still very much rely on my house staff to do that, what I call librarian work, and that's not as necessary with the technologies we have, but we haven't yet designed our technology in a way that sufficiently organizes that information, so the care teams and the patient can use it effectively. I think once, as we continue to work on that with the introduction of generative AI, I think it then maps out the next set of tasks, and that is after I gather that information, after you and I have a conversation and make a plan that works for you. I now need to feed all that information back into your data trail. And right now, what that requires me to do practically just type it all. And that is a huge amount of a time burden on current clinicians. We talked about pajama time or work outside of work time and all these other metrics which speak to the fact there's just too much to do in the time we have. And the reality is we don't need to do this. We still have, under adoption of our voice technologies, to be able to capture ambient conversations.

Thomas Maddox:
And we have under development, like just the beginnings of development, of taking all that ambient information and then putting it into a form that I use a clinical note, maybe a patient education piece for you to think about and read and refer to authorization for your insurance company to pay for the medical care that you need, all the other sort of clerical tasks that need to generate to be able to advance your care. Now, we can see, at least in some of the early signs, that with ambient voice recording, with generative AI analysis and generation of these notes, we might be able to substantially remove a lot of those tasks from a day-to-day job, to where I can focus on what thing that I can only do. And this is why I'm not worried about the robots replacing the doctors you and I, and the relationship we have, and the communication and the empathy and the understanding that I can gain about what you need and how I can help, be your guide and coach in the things to improve your health. That's not something you can outsource. And in fact, if you can remove from my day-to-day the clerical, I'll call it crap to be able to focus on that and give you the attention and the deep consideration that is needed for truly effective partnership for healthcare. That to me would be a game changer, both in job satisfaction and then also obviously in the care that we can provide. I often say the technology isn't replacing anything. That technology actually enables, the humanity of healthcare, doesn't replace it.

Steve Lieber:
Yeah, that's very insightful in recognizing what we've always talked about of operating at the highest level of your license. Take away those things you don't need to do. And there are a lot of them, as you've just articulated, that are well done by machines and such great insight as you look at the work you're doing in the Innovation Lab. And you may not be running up against this, but you're obviously reading about the policy discussions in Washington and Jefferson City or wherever. Are you sensing that you are going to need to move faster than policymakers? Will you move ahead, keeping in mind what they're talking about but can't wait? Or is there a process by which we need to go which says, okay, stick to billing and note taking, but stay out of the clinical area until we get farther down the policy and potentially regulation path. What's your read on on that intersection between policy, regulation and practice?

Thomas Maddox:
Yeah, the reality is healthcare is very much governed by both the regulations and the reimbursement policies that are in place. And people may be frustrated by the characteristics of each of those areas, but you can't operate outside of them. policymakersAt least you can't for very long. And so I think what we're going to have to do is be effective partners with our policymakers in thinking about how do we take what we're seeing as the frontiers of innovation in clinical care, digital health information management, and then advise and work with our policymakers to say what regulatory and reimbursement frameworks would best use these capabilities to do what we're all interested in. And that is providing optimal care at a reasonable cost. On the regulation side, one thing that is truly outdated and that we need to do a better job with is around privacy regulations. And this is obviously a very important area for healthcare because health data is incredibly sensitive, it can be used to ill effect. So we need to make sure that we protect it on behalf of our patients, but with digital tools, as we all know from Big Tech, that has been a bit of a Wild West. And so often what happens is the guardians of privacy and healthcare will look at the wild West of Big Tech, and they're like, that place is a mess.

Thomas Maddox:
I don't know what else to do except to say we can't use any of it because it's so messy, and that just completely robs you of the ability to leverage the innovations were seeing. At the same time, we can't allow the Wild West ethos to govern healthcare. I don't think anybody wants that either. So I am encouraged by what we're seeing coming out of Europe and a few of the other countries that are a little bit more progressive about who owns the data, how do you understand its protection, its security, and permissions behind it. Who allows who to have what permission, when? And I think as we start to clarify that on a societal level, we'll need to translate it into healthcare. Most of your listeners probably know that HIPAA was written in the mid-90s, and the technology landscape was markedly different. So it did what it needed to do to a degree then, but it is woefully outdated now and doesn't at all speak to digital privacy and what that means. I think working with policymakers on the regulation side will be really important, particularly on the privacy side, with digital data on the reimbursement side of the move to value-based care, does appear to be progressing, albeit slower than anybody would like. We do know that most of the digital tools, because they are providing predictive insight that they're allowing you to ideally head off medical conditions or at least minimize their complications. That kind of runs at cross purposes of how a lot of healthcare is paid for, because healthcare right now is largely paid, as the more I do to you, the more I get paid. And if you're not as sick, there's less for me to do, and so I get paid less. And I don't think anybody is saying, oh, please continue to get sick so I can make money. But incentives matter. And the fact of the matter is, the system moves everybody in that direction, whether we like it or not. So I think continuing to encourage, I'm actually encouraged by what I see some of the big commercial payers doing where they're setting up more value-based contracts with healthcare systems like ours and say, hey, we think this person will cost as much over the next year, we'll just agree to pay that you manage the best you can and will reward you if they come away pretty healthy at the end of the year. That really starts to align us in saying, okay, so now I do need to invest in some of these tools and information and digital technologies to really keep you as healthy as long as possible, because that is the most cost-effective thing for me to do. In addition to obviously being the right thing to do for you and why you came to us. So I think when we see when we do the ROI of our various innovation pilots, if we study it in an at-risk population, in our accountable care organization, or with an employer who owns the total cost of care of their employees, the ROI is very high. But we also see that just because there is so much volume in our health system and everything we can do to be economical and the care that we provide you, even under a fee-for-service structure, even then, the digital tools often have a smaller but still positive ROI. So I think we can move in both reimbursement schemas, but I think both to maximize the value of these digital tools. And frankly, the best thing for the patient is to as quickly as we can get to more of that value-based framework and let that be a really complimentary incentive to some of the innovations we're seeing in our lab.

Steve Lieber:
Yeah, makes a lot of sense. I want to circle back to one thing you said early on about community-based care versus academic medical centers. We certainly look back into 15 or so years ago. —theThe early adopters of EMS generally came out of medical and academic medical centers. When think about the hems in REM stage seven, the first ones were almost always academic medical centers. The investment was the endowments and the money was there. The research orientation. If we can think back that far, it was new and fell into that R&D category. Our academic medical centers leading the effort again for some of the same reasons in the adoption of these newer generative, and let's just stick with the ubiquitous term right now, for lack of any other, because there are a bunch of other tools as well that might fit in that really out there, innovative category folks like you, where we ought to be looking to see the early findings because it's going to go both ways. Works doesn't work so well, needs to be worked on some more.

Thomas Maddox:
The answer, I think, is the nuance I would, so I would say in general, at least if I understand your question correctly, the leading digital health medical systems in our country are not academic medical centers. And in fact, I think there's some of these bigger health systems that both have an operational priority. And frankly, the resources to really be able to invest and roll out and operationalize a lot of these digital capabilities. I've seen really good things coming out of Ascension. They've been doing a lot of progressive things. Providence out on the West Coast has been doing some really interesting things. Atrium in North Carolina, has just been leading a lot of good things, and even some of our four prophets was speaking to some of the innovation folks at HCA recently, and they're just doing really it's really amazing things, often driven by the fact that there are enormous systems. And if they don't manage well, nobody's operating with huge margins. So if they don't operate well, they're going to get underwater really quickly. That said, it is true that true R&D budgets often still live in academic medical centers, largely in the research arm of the School of Medicine and largely fueled by the grants that come from NIH and other big funders, often the government. The disconnect, but I'm hopeful this will change over time is that often in the research world, your incentive is not to identify new things that then translate to operations. Your incentive is to identify new things and then publish a paper about it and go get a new grant, back to incentives. Incentives matter. So what ends up is you having this sort of isolated cycle of R&D. It's really more R than D. And so it's all these research insights that don't often make it across the street into the healthcare operations.

Thomas Maddox:
And one thing that I've tried to help in our lab, and I've seen peers do this at other AMCs around the country, is can we serve as an application bridge between some of the insights that our researchers are doing and then the operations that the bulk of our employees and our healthcare system are working on every day. And we're lucky that we have a 14-hospital system. Two of them are teaching hospitals and the 12 are community based hospitals. So in some ways, we have the right lab. And what I found in the hot term in AMC is right now is a learning healthcare system, because I think we have the right raw material for a true learning healthcare system that our lab can say, ooh, this is an interesting idea coming out of our researchers. Let's now start to pilot, test, and scale it, see how it works in the wild, in the actual day-to-day healthcare operations, and bring the research methodology in this iterative, sprint-based fashion of collecting the data on how it's working, analyzing it, and feeding those insights back to our research groups. To say to your point, this part worked. This part didn't. Based on that new information, what can we learn and what's version 2.0 and then start setting up this virtuous R&D cycle? And in some ways, I think AMCs are primed to do that. But I will say that it's going to take new mindsets. And sometimes the leaders of AMCs just because they've built their career on the old research paradigm that I've described, it's a little hard for them to flip the model. And so I think that may be one reason why they're lagging a little bit.

Steve Lieber:
Excellent. Great insights. So to wrap up, our listeners or CIOs, CMIOs other clinical leaders, digital health leaders, and their teams, what's your takeaway? What's the one piece to share with the audience as we wrap up here?

Thomas Maddox:
I think the thing that we have gotten religion on, and that I really think is probably a promising approach to realizing that learning healthcare system and seeing where digital technologies can help us out in the quickest and most informed way possible, is that we really adopted the so-called product model that we've learned from Big Tech. And so we have invested in user-centered design, where we get really close to our patients and care teams, conduct ethnographic research, really understand the problems they're trying to solve and the jobs that they need to do. And then we have product teams that are cross-functional teams and use product management philosophy of identifying what are the highest value things we can test and put into production, and see if they work to improve care and do it using some of the agile techniques that software companies use. Coming up with small proofs of concept two week sprints, rigorous collection of data and analysis of that data, and an ongoing cycle of learning to be able to very quickly speak to what the user and the health system needs to deliver healthcare and do it sometimes in a matter of weeks, if not months, rather than the traditional years long time frame. So we've just started moving into that arena. Like I mentioned, we have product teams, we have designers, and we're already starting to see some really good impacts with our primary care patients, as well as with some of our nurses, and bringing digital tools in a really rapid way, and the data proven way that improves their experience and the outcomes. So I would encourage my fellow technologists in healthcare to see if this product model is something that we should continue to bring into healthcare and realize some of the value that it can provide for our patients.

Steve Lieber:
Excellent. Really appreciate that. We have covered a lot of territory in a short period of time, and I really do appreciate the time that you've given us today, Dr. Maddox.

Thomas Maddox:
Now, it's my pleasure. I love doing this stuff, and it's just great that the audience you have are fellow journeymen and women in the software. Excellent.

Steve Lieber:
Thank you. And to our listeners, thank you for joining us today. I hope this series helps you make health care smarter and move at the speed of tech. Be well.

Intro/Outro:
Thanks for listening to Smart From the Start. For best practices in AI, in ambient intelligence, and ways your organization can help lead the era of smart hospitals, visit us at smarthospital.ai. And for information on the leading smart care facility platform, visit care.ai.

Sonix is the world’s most advanced automated transcription, translation, and subtitling platform. Fast, accurate, and affordable.

Automatically convert your mp3 files to text (txt file), Microsoft Word (docx file), and SubRip Subtitle (srt file) in minutes.

Sonix has many features that you'd love including collaboration tools, automated subtitles, automated translation, advanced search, and easily transcribe your Zoom meetings. Try Sonix for free today.