Episode 7 : Tech, Trust, and Transformation:

Smart from the Start_Tarun Kapoor: Audio automatically transcribed by Sonix

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

Smart from the Start 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 HIMSS, as he interviews the brightest minds in the health provider 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 is 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. Tarun Kapoor. Dr. Kapoor is Senior Vice President and Chief Digital Transformation Officer at Virtua Health. In this role, he oversees Virtua's Digital Transformation Office and orchestrates Virtua's enterprise-wide master plan in support of an intuitive care journey for all consumers. Previously, Dr. Kapoor was the president of Virtua Physicians Partners and the Senior Vice President and Chief Medical Officer for Virtua Medical Group. Prior to joining Virtua, Dr. Kapoor was Regional Director for EmCare's Mid-Atlantic Hospitalist Division. Welcome, Dr. Kapoor.

Tarun Kapoor:
Hi, Steve. Thank you so much for that very kind introduction.

Steve Lieber:
You're quite welcome. And it is a pleasure to be with you today and have a conversation with a physician who has a significant role in leading digital transformation. We've got a couple of intersections there, some disciplines, technology and medicine, and let's start out there. My career goes back to the early days of EHRs, and we had a lot of challenges getting clinicians to buy into what was going on. In that period of time, we may have moved one generation in terms of the clinicians practicing today, I don't know, there's still some of us still around, but kind of take us through, short, sort of the evolution, hopefully, there has been one, around physician opinion and reaction to technology in general.

Tarun Kapoor:
I know that the rub sometimes is, for physicians, or I would say, clinicians in general, including anyone who provides front-level care, whether it be our advanced practitioners, our nurses, etc., to be a little bit skeptical about technology, and their adoption curves within the clinical realm, just like they are anywhere else, right? You got your early adopters, and you have your laggards. I think what sometimes throws the clinical field off on this is, yeah, well, I see the promise of the technology on paper, but then the actuality of what I'm dealing with at 2 a.m. in the, you know, in the middle of the night is something completely different than what's on a PowerPoint slide. And, you know, you see a lot of great PowerPoint slides, a lot of times the execution is a little bit questionable. I'll give an example. And I'm an informaticist, so I, by training, and so I know what it's like to talk to doctors who are skeptical on this. But I remember, once I, it was probably late at night, I was in the hospital meeting somebody, and they say, Hey, look, with our consolidated EMR, you can now see all the previous discharge summaries this patient has ever had for the last 15 years. And on PowerPoint, that's wonderful, but when I'm sitting there trying to admit somebody, and I still have the exact same amount of time to admit them, because I still have 12 other people who are waiting to get admitted, and I realize I don't have time to read 15 years of ... and discharge summaries, I think that's where we hit the nadir of this, where we had all of this consolidated information, but what we didn't have was processed knowledge. And perhaps, what I think folks are on the technology and also in the clinical side are most excited about right now, is that now we are in a place where we can go from that nadir where just volumes of data to say, Hey, can you process this and sum it up for me? And has this person ever had this in the last 15 years? And ask the machine to look at that without me having to read 15 years of medical records? That's what I'm very hopeful for, but it has been a journey and sometimes quite a painful one.

Steve Lieber:
Yeah, I want to get to that because you really have carried us to where we're hoping we're on perhaps the cusp of a next generation. So last generation was about digitizing information. We didn't have it in a digitized form before, we were trying to get off paper, it was all about data collection. I mean, it shouldn't have been, but it was. And now, as you said, something to work with that pulls information out that I need right now that helps me with the next thing I need to do. And so that leads us, obviously, into the two words of the day, artificial intelligence. And so clearly, the intent there is to be able to comb through large amounts of data and information and be able to extract actionable activity out of that. Where are you in terms of virtuous work in this area, and do you have a sense of clinician reaction to this next generation of technology that's being thrown at them?

Tarun Kapoor:
We have some examples here where we think we're making some good progress, where I think it is, there are lots of words thrown out there. One day, I'm hoping I'll be famous for something, and so I put out something called Kapoor's Inverse Law of AI, and that is the number of times a person mentions AI in their presentations, inversely proportional to their understanding of AI. So I will try to keep my usage of AI to a minimum, but what I do find is a lot of times we're we're tossing the word AI and automation back and forth interchangeably, and they're not. Where I think there is an enormous opportunity right now is with automation. And clinicians feel way more comfortable around automation, where the human rights, the rules, and the machine processes within the construct of the rules. Whereas you could say to the, maybe further to the other extreme of AI is that we allow the machine to rewrite the rules as it goes on and learns. I think we're way more uncomfortable with that right now, and as well as we should be, having, letting the machine rewrite the rules. So let's start from a point of comfort, and I think the term you mentioned is absolutely spot-on, fear, right? Because look at how we were trained. We were trained in gladiatorial fashion, that if you did anywhere from 3 to 7 years of, you survived gladiator school, which is known as your residency. At any rate, you're going to see this patient at 2 a.m., and you're going to be 36 hours sleep-deprived, but you're going to still do it the right way because we've beaten it into you to do it the right way. Now, all we're going to say, I have a machine to ... me, don't worry about doing it this old way. I've got a new way, I figured it out on my own, without really any 100% confidence that I know for certain that this thing didn't hallucinate on ..., and it just made up its own rules. So let's work from an automation point of view, and I can give you an example that has had immediate impact. So one of the things that we work on, ejection fraction for the non-clinical listeners, is the amount of blood that your heart squeezes out with each beat, and you want your ejection fraction to be above 35%. Well, when we went to look in our charts, like where is the ejection fraction on all these different patients? It's all over the place, right? Sometimes it's in a discrete field, sometimes it's in progress note. Well, if we wanted to do a chart review on these 6 million progress notes, it would have taken us, Steve, 27 years to pull out all of that information. Our advanced analytics team came out with an algorithm, it read it in a weekend. Now we found all of these opportunities to treat folks who are not getting optimal care, we think, to reach out to them and say, Hey, listen, we think we can do better for you. That's how you get to the heart of the clinician. It's a meaningful, something I can do with the data, not just the data for the sake of the data, something that we came up with that was actionable and that we're having that conversation with those patients now.

Steve Lieber:
When you look at automation, or even you mentioned data analytics and all, where is your philosophy in terms of, go slow, the cost of status quo and incrementalism versus a more aggressive approach in terms of digital transformation? What are you saying to your teams and to your executive suite about how you should proceed?

Tarun Kapoor:
Everyone wants everything to be transformative, and the answer is, that's, by definition, it can't. Not everything can be transformative, and nor should everything be transformative. So if you imagine a two by two where the one axis is degree of impact or transformative capability and then the other axis is feasibility of pulling it off, low feasibility, low transfer, low impact, everyone knows to stay away from that. That's not hard. Highly transformative, highly transfer, highly feasible, everyone knows to go after that. It's the two other boxes where there's a mismatch. And what we say is, it's okay to be evolutionary in some things because that is still progress. But what ends up happening is, as you're continuing to evolve, then you start to realize, wait a minute, we now got to a place that we went from maybe high feasible but low impact to, now it's in the upper corner. So things that were previously evolutionary can turn into transformative things. Not everything is transformative on day one. The best real-world example I can think of, and my team's sick and tired of me talking about this, is Netflix. Netflix was not transformative on day one. It wasn't, right? They mailed DVDs to your home, and there was a delay in getting them. Oh, and then they came out with their streaming service, and if anyone remembers Netflix's streaming service to begin with, it was awful. It was just like every single foreign documentary you never wanted to watch, right? But they kept at it, they kept on it, and then eventually, maybe 4 or 5 cycles into it, they realized, oh, the transformative piece is that we got to write our own content. And then, when Netflix started producing everything themselves and put that onto their platform, that's when the transformation truly happened. Evolution is not a bad thing. However, once in a while, all the circles line up for transformation to occur, and what we look for is we look for three things to line up. We look for, is the technology ready to do this? Is the patient consumer ready for the transformation? And they're excited about it. And then the third one, is the operator ready to do that? And then when all three of those circles line up, that's magic time, and you go as fast as you can, right? That's transformative steroid time. And there are some unique things that are starting to happen right now that we're seeing all three of those things line up. I think the other example that happened during COVID was telemedicine. Technology was ready for decades, but the consumer wasn't ready, and the operator wasn't ready. And then, all lo and behold, they all aligned. Interestingly, some of those circles are starting to disengage now post-COVID, right? And so that's not...

Steve Lieber:
Particularly around telemedicine, specifically.

Tarun Kapoor:
Technology is still ready, but some of our patients are saying, I want to come back in person, right? And some of our patients and some of our doctors, I want to see you in person. So it's not a static thing where it's always happening, it's constantly ebbing and flowing.

Steve Lieber:
Is there, because you said, having all three of those aligned at the same time is magic, which it's, I'll translate that into, doesn't happen all the time.

Tarun Kapoor:
Very rarely.

Steve Lieber:
Is there one of those drivers that maybe when it's ready, and the other two are close, yeah, that's a go time? I mean, do you have a hierarchy there in terms of where you look?

Tarun Kapoor:
Yeah, so I would say our rule of thumb here is if one of the three circles, only one of the three circles is ready, no go. Set, let it sit back, we'll keep watching it. If two out of the three circles are aligned, then you got to think through, okay, what is it going to do to take the third one to get in? So let's say the technology is ready, and the operator is ready, and I think those are two of the harder ones, two hard ones to control, then you got to go out and educate the consumer on how to adopt that. That can be pretty hard to do. But if the consumer is ready and the operator is ready, then that's where I'm going out and looking for startups and some cutting-edge technologies. Hey, hey, listen, we want to do this, and look, this is, alignment has worked out beautifully, but I need you to kind of co-develop with us, is how we think about it. And then the last one, in which I do think is probably, well, very hard, but to some extent controllable, is getting the operator to change. And I think you asked me what is a key takeaway, just a heads up on the takeaway is, it's all about the execution, and it's the adoption from the operator where I spend, my team spends, the vast majority of our time.

Steve Lieber:
We learned that 15, 20 years ago, that you can't just throw technology in and not change the workflow, not pay attention to the way clinicians practice, and perhaps work on changing that workflow. Still the case? I mean, we, is that still a key piece of an adoption and new installation?

Tarun Kapoor:
I would say more so than ever because you still have to deal with the facts of all the other times we got it wrong before, and people still remember. So there's still hesitancy. It's like, Oh, wait a minute, this was supposed to make things better, and now we're coming back the fourth time. This time is different. Well, this time better be different, and we're going to have to work through that. I do think this time is different, but if we've made a lot of mistakes in the past, and I think we have some work to do in terms of building; because the opposite of fear is trust, and I think one of the things you talk about is working at the speed of technology that is going to be completely pegged, though, to the speed of trust. They go hand in hand.

Steve Lieber:
Yeah, that's really key, and that trust, in some of the conversations we've had in this series, that question of an issue of trust has come up, and it's caused me to really think about the past. Because we've, yeah, there are a lot of new people in the field now that weren't there 15 years ago, but there are a lot of us that are, and building that trust, and it's, okay, Did you learn from the first time? We went through something like this, and such is really key. I think that's a great insight. There are, and I like your point about the number of times you mentioned AI is inversely related to how much someone knows about it, but as you're looking for new technologies and all, how are you filtering through that? Because you're right. Right now, I think you can probably look at you go to anybody's trade show in healthcare technology, and AI is on everybody's backdrop, whether it really is or not, it's on the backdrop. How do you filter through that?

Tarun Kapoor:
Tell you, our approach, I can't necessarily say it's right for everyone, is there are times in places for point solutions, right? I have an immediate need, and I need to fix it now, and fine, I'm fine with this point solution helping that. The problem that I think a lot of health systems, health delivery systems are going to run into in the upcoming years is these point solutions start to add up really fast in terms of cost perspective, right? There's, these are all six-digit, sometimes seven-digit point solutions, and we have literally hundreds, if not thousands of use cases. So do the multiplication, and I'm going to tell you right now, if you look at the margins for the majority of health systems in this country, you cannot support this. All of our incremental operating margin, I think the average operating margin across the United States for a not-for-profit health system is around 1%. All of our operating incremental or all of our operating margin between now and the rest of this decade could be all consumed by point solutions' AI, it's just not feasible. So then what we're trying to do is we're trying to find, where are few platforms that we can work off and build and iterate such that each incremental use case is not costing the exact same amount as the last one, right? So I'm actually getting some type of still getting returned, but I'm still starting, getting some gain in efficiency in the cost structure. Not that we're not open to experimenting with point solutions, but we're going to have to continue to do this because there are such enormous constraints from a cost perspective, and we're just not going to be able to continue to crank prices up on the services we deliver. The other piece that is just an unescapable fact is, from a human resources perspective, we are not going to be able to hire our way out of some of these problems. It's mathematically impossible. And so therefore, we're going to have to be a little bit more creative with what we're able to adopt.

Steve Lieber:
Some of the conversations we've had have been around patient monitoring as an example of those point solutions versus platforms. The camera is great, but if all it does is point at the patient in the bed, or whatever, and not anything else in terms of other functionalities built into a platform. You are, you're in a sense, having to buy multiple-point solutions instead of one platform to address a range of activities that go on in the patient's room.

Tarun Kapoor:
Yeah, and I mean, I think that's how we're thinking about it as well is, Okay, great, instead of having a one-to-one sitter with someone in the room now I have maybe a person can watch 12 feeds, right? Okay, but you know what? That's still, now we're at the maximum of that point, right? There's that asymptote we've hit. But if I can tee up and watch, but why only 12 people get to be watched, right? Why can't I watch the whole house? Doesn't everyone deserve to be monitored for a fall? And so that's where we're trying to look for scale stuff. Now, if I start to see something that's worrisome, I don't mind escalating to a human and then to turn on ...

Steve Lieber:
Which kind of is in the purpose of monitoring across scale and then having something give you an alert so that you can send that person to the place where they need to be instead of wandering around, maybe crossing paths at the right time.

Tarun Kapoor:
Gosh, it probably goes back to, I think it was, I forget the brothers who did The Matrix, right? But what was the quote? Never send a human to do a machine's job. But there is plenty of work for us humans to do.

Steve Lieber:
Oh, absolutely, because it works the reverse as well. There are times where you need that sort of additional observation to see, because there can be misinformation crosswires, you know, that sort of, and so there is a time where a machine only can carry it so far.

Tarun Kapoor:
Yes. And I think that's what we're really excited about right now, as we're seeing a few opportunities to be truly transformational right from the get-go on this. Yes, in some cases, we will continue to iterate, but the idea of ambient listening is what I think gives me a lot of hope for what we can do. I mean, the other reality of it is our clinical staff is exhausted. There are already pretty significant signs of burnout amongst clinicians, both from physicians, nurses, etc., before the pandemic. Coming out of the pandemic, we've gone from big problem to flat-out crisis, and it's universal, and it's not just in the United States. So it's not like, hey, again, we're not going to hire our way out of this. We're not going to import our way out of this either. Everyone's struggling with this across the world. So we're going to have to come up with other solutions, and I think going through that triple circle piece, the technology is, well, there is some stuff that's really happening nicely with technology now. Our clinicians are like, we need whatever help we can get, and then I think the patient consumers are, maybe there's a, we don't think our patient consumers are like, wait a minute. No, I'm okay with a set of eyes helping me make sure I'm safe. And I think that is where maybe all three circles aren't perfectly unified yet, but they're starting to converge pretty nicely, and that's where we're making a pretty big bet.

Steve Lieber:
I like that in terms of the label you put on that ambient listening because there are multiple pieces to that you're going to extract from a platform type of solution there to be able to guide clinicians. And as a consumer, explain it to me as to why what it's going to do for me, and it's like, I'm there. Because it really does, it adds to a degree of comfort that even though someone's not here, you're listening, you're looking out for me, you're trying to make sure that the best possible things can happen here.

Tarun Kapoor:
If you think about it, we have these technologies in our homes, and we don't even think about it, right? Whether you have a Nest or an Ecobee or whatever, your smart thermostat, it knows whether you're in the room or not. How does, it knows because that's what the sensor is supposed to tell you, that ambient sensor says this person is in the room. If you have a security system, you have a motion sensor. The motion sensor can tell the difference of whether you're a human or a dog, because that's how it knows not to set itself off if you set it on alarm, but it's pet sensitive. So these things are all around us, and we've entrusted our safety to it in our home settings. Well, a healthcare setting is actually a pretty complex place where we could all benefit from a little extra watching and guiding as well. So I think that's again, where we can get very excited about. I think we have to be mindful and be very thoughtful from some of the other angles that people could be concerned about and no, this is not replacing your bedside nurse. This is augmenting, that's the word ...

Steve Lieber:
Augmenting.

Tarun Kapoor:
We are augmenting the skill set that's around you so that when we do spend time with you, we're not being distracted by an alert call going off here and oh, sorry. I'll be back in five minutes. Another alert goes off, I'll be back in five minutes. When I'm with you, I'm with you. And I think that's what is very helpful for both our patients and as well as our clinicians.

Steve Lieber:
And that's what the clinicians have been asking for, from the first time we started bringing information technology into the bedside is, let me spend time with the patient. Don't put the technology between me and the patient. You touched on it earlier, but let's wrap it up with your key takeaway for CIOs, CMIOs, CNIOs, the people that are bridging the gap between technology and clinical care at one end, the other end, and the middle. What's the key takeaway you've got to leave them with?

Tarun Kapoor:
Our team, we use a formula that we're not exactly sure where it came from, someone stole it from someone, who may have stolen it from one of your old colleagues, Halle Wolf. And that is the formula, NT plus OO equals COO. New technology plus old organization equals cost the old organization. This is not about the NT, this is all about the OO, and the willingness and the courage to change the OO. And our team, what we say and think it's a phrase that we embrace, it's about cold-blooded execution. Everyone's going to have these tools available to them, the price points on these tools will continue to drop as they democratize, but it's those folks who are able to operationalize them with just absolute precision and focus and energy and urgency that are going to see the day through. I think for the first time in healthcare delivery, like it or not, there's going to be winners and losers. That has never happened before. In the private world, that happens every day. I think it is a brave new world, but I do feel comforted that those who are willing to be bold, but most importantly, boldly execute, are going to have a tremendous advantage.

Steve Lieber:
Excellent. Tarun, it's great to see you and chat with you. I sure do appreciate you being with us today.

Tarun Kapoor:
Steve, many, many, many thanks to you and the team for the opportunity to share some ideas.

Steve Lieber:
Great, thank you. And to our listeners, thank you for joining us. I hope this series helps you make healthcare smarter and move with the speed of tech. Be well.

Smart from the Start Intro/Outro:
Thanks for listening to Smart from the Start. For best practices in AI and 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 enterprise-grade admin tools, world-class support, collaboration tools, automatic transcription software, and easily transcribe your Zoom meetings. Try Sonix for free today.