July 6: Today on TownHall is the first of two parts of a conversation between Brett Oliver, Family Physician and Chief Medical Information Officer at Baptist Health and John League, Managing Director, Digital Health Research at Advisory Board. In the discussion of AI implementation in healthcare, how do the experiences with other technologies inform the expectations? Is AI's path unique or following a familiar hype curve? How does digitizing and automating patient intake processes affect patient satisfaction and their overall healthcare experience? What challenges might organizations face while implementing AI-driven intake solutions? Considering the promise of AI in streamlining processes like ambient documentation, how do we reconcile the potential benefits with the fears and uncertainties? What do the experiences with existing processes reveal about the need for AI in healthcare?
"The Patient Experience - A Technology Perspective" is a live webinar that explores the intersection of healthcare and technology, focusing on enhancing the patient experience. As healthcare systems prioritize patient-centered care, leveraging technology becomes crucial. Join us on July 6th, 1:00 PM ET and join the discussion! Register Here. - https://thisweekhealth.com/leader-series-the-patient-experience-a-technology-perspective/
This transcription is provided by artificial intelligence. We believe in technology but understand that even the smartest robots can sometimes get speech recognition wrong.
Today on This Week Health.
what we could do with chat GPT is take all of that and summarize. Here's the top two recommendations during this visit.
That to me is. Scary, but also isn't that what physicians have been asking for? They've been deluged for years by a ton of data feeds, but no real insight
Welcome to TownHall. A show hosted by leaders on the front lines with interviews of people making things happen in healthcare with technology. My name is Bill Russell, the creator of This Week Health, a set of channels dedicated to keeping health IT staff and engaged. For five years we've been making podcasts that amplify great thinking to propel healthcare forward. We want to thank our show partners, MEDITECH and Transcarent, for investing in our mission to develop the next generation of health leaders now onto our show.
Hello and welcome. I'm Brett Oliver, the CMIO for Baptist Health in Kentucky and Southern Indiana, and I'm tickled to death to have back John league from the advisory board, John's the managing director there at the advisory board. And we're going to talk a little bit today about AI and how technology can support our workforces and healthcare.
So welcome back, John. It's been too long.
Thanks, Brett. It's great to be with you again.
Yeah. I appreciate you taking the time. Yeah. So let's jump right in. From my perspective, imaging, AI and computer vision seems to have really taken kind of front and center for a lot of organizations that are starting to incorporate more than just a few algorithms from their EHR, a lot of activities occurring.
So, but from your perspective, you get a broad view of the industry. What's in second place? What's next? What are some areas that have really made up a lot of ground that maybe we're not thinking about as healthcare organizations?
I think the number two when we think about how we see AI is basically anything transactional, how can we automate this process?
How can we consolidate a lot of the steps and remove that manual burden? I think a lot of that is well intentioned and in the right place. And certainly I think when we look at how much organizations she's Have invested in serving a lot of those sort of transactional processes, whether that's rev cycle or med rec or something like that.
I think all of those are very valuable, but when we talk about how should we be using AI, even when we think about transactional stuff, I think we're often skipping over the question of. Should we even be doing this? Are we simply automating a bad process? Are we making something we don't want happen more efficiently?
So we're getting more of what we don't want faster. Like that is, the road to nowhere. And I worry that when we focus exclusively on how Quickly. Can we eliminate the manual? How fast can we get to the end of that transaction chain? We miss out on, Hey, maybe we shouldn't be doing this in the first place.
Like just because we were doing it manually, maybe that was not the right way to do this. The value of, AI and the value of the moment I think is that we should be able to ask questions about. Is this really what we want? And the technology is such that in many cases, we can figure out maybe a new pathway.
And I think that that is the next step, I think, in a lot of ways for a lot of these technologies that are incorporating AI at any sort of sophistication of AI, whether that's just a simple algorithm or whether that's something extraordinary like chat GPT, which we'll talk about in a little while.
Yeah, I love that thought because you're exactly right. We get so busy in the day to day and oh, here's this product. Hey, that would be great. We could cut our efficiency by 50% or improve our efficiency by 50%. But maybe we ought to have that conversation on the front end before we get to that point. Or do we even need to do this or do a different way?
And I think that it's really poignant for me because I think the way we deliver care is going to change so much Over the next several years or should I don't my personal opinion is we're not going to out recruit. We're not going to be able to recruit enough nurses and enough physicians enough providers, the way things are going right now.
That's not to undermine my organization's recruiting efforts or anybody else's like we still want to do that but We have to think differently. How can we deliver care that's, better, not just maybe more efficient the way we're doing it now. So I love that. Are there any specific areas that you've seen that have worked well for organizations that have said, okay, this is a transactional approach.
This AI can, improve things make it more efficient, but we're not going to do it that way anymore.
I think a lot of the. I'm going to say the biggest gains are in things like intake revisit. Whatever. And I know that in a lot of places, folks don't think of that as, it's like, well, maybe there's a little ai, but it's not really that important.
But when you think about things like, how do you really digitize that intake of a patient when they come into clinic, like, what are the things that we usually do? Okay, Brett, the, the thing that I hate to see most when I go into a facility is, the coffee cup full of pens that have flowers taped to them.
Do you know what I mean? So you don't walk away with the pen. And the reason I hate to see that is because I know that what that coffee cup means is that I'm going to have to fill out a bunch of forms. They, that's what they're going to do for me. I think when we think about things like the, the rest of that journey, like showing them your insurance card and stuff like that, what do we usually do with that?
Well, you give that to the MA or whoever's at the front and they make a photocopy or something, right? The value of a lot of the intake solutions is that you can do all of to do everything there and using, Computer vision to, to read an insurance card and find your number and your group number.
And I don't have to look at it and go, okay, first, I got to move my glasses to make sure I can read the number. And then like, which one is the group number? Which one is my number? Like all of those things. It can just do that because it's been taught how to look at all of those things and drop it in and it'll be completely ready.
And then the only thing that I have to do manually is You know, the whoever is rooming me is going to need to just confirm like, am I john league? Is this right? Do I have this? Those kinds of things as opposed to, taking up all of that time. It shortens the weight. It makes satisfaction better.
It takes away a lot of that clerical work, both from the staff. And from the physician and a lot of those systems are actually even able to queue up very valuable information. One of the things that I have been honestly surprised by is that those kinds of systems are very popular among.
The highest productivity physicians that we have because they look at it and they're like, yes, absolutely. Why would I spend time talking to you about your allergies when you're here for CHF? Right? Like I don't, need to cover all that stuff. This is helping me add unique value that I, as this high performing, probably very proud physician am able to bring to this interaction.
I can, we can focus exactly on what you need and we can spend whatever amount of time we have. On delivering as much value in that time as possible. I think that alone in terms of like satisfaction and productivity is enormously valuable and just shows that if you can solve for those little transactional pieces with a different, with a slightly different process, you can also engineer benefits way downstream.
Oh, a hundred percent. I love that because it's not just that administrative burden on your front office staff. You kind of bled into that. Then I'm in the exam room or I'm in the hospital room and the AI brings those important things together that I'm not having to go flip through your chart with the chart on the computer, ask you whatever it is.
We can get down to what is that unique value that I can bring to the moment right there, right away. Yeah. I love that. Yeah. Well, we mentioned it before we'll switch gears here just a little bit, but not really, I suppose with some generative AI. I mean, it seems like it's moving. At breakneck speed, or at least the interest is moving at breakneck speed.
Just general thoughts here. I know we're kind of early in this journey, but is it the next bright and shiny thing, or do you think it's got a real chance to be transformative?
I think it does have a chance to be transformative. And I say that as someone who, at the beginning of November. Was actively planning not to talk about AI this year in any of the work that we're doing.
It's like, Oh, okay, fine. That, you know, I mean, there's advancements. It'll be fine. And then chat GPT. I'm like, okay, this is an interesting little toy. As we've seen the ways that its uses have evolved and its capabilities have, grown sort of exponentially off of pretty much the same base of technology.
] I'm really very bullish on that technology. One of the things that we hear when we talk to vendors to big tech companies is even they are surprised at how fast the uses of the technology are evolving. And I think when you look at, you know, there was, a news. Last week about how several AI leaders and Elon Musk was among them were like, we need to have a moratorium on this.
We need to need to have a holding period. I don't think that's necessarily because they're afraid that AI is going to take over the world. I think it's because they can't keep up that it's evolving at a pace that they can't even figure out where it's going to go next. And I think. That speaks to the power of the technology, and also, like, we need to be thinking about how this is going to manifest in healthcare.
Because it is going to come... sooner rather than later. I already know physicians , who use it not in clinic, but who use it personally, for all kinds of needs, , the consumer adoption element of it, the adoption. And remember that. Healthcare workers, clinicians are also consumers outside of the healthcare organization like this is so this is being so broadly used that we have to figure out how to deal with the things that folks are bringing to our organizations.
Well, hey, I asked chat GPT if I had cancer and here's what it's, you know, I mean, things like that that we used to do with. Dr. Google with WebMD, folks are going to chat GPT and getting potentially pretty comprehensive, non diagnostic clearly answers that we will need to understand how to deal with, how to approach, how to respond to.
And I, I don't think, I don't think we're ready. I think we've ] been as an industry sort of by previous enthusiasms for AI. We did a survey of clinical leaders last year at advisory board and asked them sort of how they felt about AI. And we had done the same survey in 2018 and 2018 was about 70% of people who were positive.
So they answered either AI will be transformative. Or AI will generate incremental value and it was about 45, 25, maybe break down transformative incremental same survey last year, 70% again, still positive, so it hadn't grown, but it also hadn't declined, but the, the internal was flipped. It was 45% incremental and 25% transformational.
I really want to run that survey again, very soon to see how chat GPT has pushed that in one way or another.
📍 📍 We'll get back to our show in just a minute. Ever wonder how technology can reshape the patient experience? Join us for our next live webinar, the Patient Experience, a Technology Perspective on July 6th. At 1:00 PM Eastern Time, we're bringing together expert speakers to dive into the intersection of technology and healthcare.
We're gonna explore topics on digital health tools and the impact of ai, blockchain, and other things around this. Whether you're a CIO or part of a healthcare IT team, we think you will gain practical insights from this discussion. Uh, don't miss out on this conversation. Register today at this week, health.com.
We hope to see you there. Now, back to our show.
Yeah, that makes a lot of sense. I guess that's a natural. Have you found that with other technologies historically that, you get on the hype curve and it's like, this is going to be transformation and then you actually get into it and realize, okay, this is going to be good, but it's going to be an iterative approach.
Or was that relatively unique to AI back when you did the survey?
That's, that's a great question. I think the challenge that healthcare has with AI that is different from some of the other technologies that we have incorporated is that AI has been wildly successful in other industries for those sorts of applications.
And there's not really a There's not really a slow down and wait and see version of it in a lot of ways, when we think about the adoption of other technologies I think each one has its moment, but I think we always assume that over time, we will be incremental with it, whether that is some sort of actual Procedural sort of thing, some sort of actual treatment, whether it's a surgery, something like that whether that is, say telehealth.
I mean, if you, sort of ignore the giant bump from covet, I mean, telehealth has been on a pretty clear path, glide path over time, iterating, figuring out what that is I think this is unique because it is. So prevalent in other industries and by its nature, it tries to get outside its own box.
I'm not sure that we have a good way to understand how to use that in healthcare. Yeah,
that's fair. For those that you feel like have a decent handle on a process, or maybe there aren't any, I'm curious at a broader look at the industry, how are organizations judging , these new offerings, evaluating them?
Specifically with the way we've seen generative AI, everyone is very Interested, but hands off at this point, like we would love to see a way to do this, but there is a, there's a lot of concern about safety and there always should be right. Like, that is, that is totally, that is totally understandable.
I think the interesting thing about a potentially transformative technology. Like, this is probably the way to approach it is to think more about what are the big problems. And how does, a use of this technology work towards addressing that? So one of the things that, we've been thinking a lot about is ambient documentation.
So nuance or a bridge or whatever, whatever your platform is. In thinking about, about how that works, if you think about adding chat GPT into something like that, and you layer that on with. Increasingly flexible interoperability, there's a world where your, user interface interface for the physician could be something like, okay, so we're interoperable.
So the record is able to pull in not only all of the data from our system, but maybe data from an HIE, maybe data from a payer that we are working with. And then, like, all of that stuff is stuff we can do today if we have the right relationships. Because we've got the APIs and all of that stuff actually does work.
But what we could do with chat GPT is take all of that and summarize. You see it in one view. And if it can be safely trained, we could also get, you know, here's a summary of what's happening here. Our, our top two recommend here's the top two recommendations for what you need to investigate during this visit.
That to me is. Scary, but also isn't that what physicians have been asking for? They've been deluged for years by a ton of data feeds, but no real insight, right? This has the potential to get. Now, again, as I said, my number one thing was it's kind of scary. There's a lot to work out before we get to that point.
But if you're thinking about like, what are the big challenges? That's sort of the way I think that this kind of technology. naturally lends itself to in a very exciting way, a very scary way again. But, I think , somebody is going to be doing this. We need to, we need to be figuring out , what is the path that we want to get there?
No, I love that.
Cause then in the inpatient space, you think about nurse, how much time does the nurse spend at the end of the shift, checking out what's happened? Here's what's going on next. How about a handoff from one hospital to the next? And then it becomes static. Here's my, patient list.
And I think I read a study, it's been a while, but I want to say that after two and a half to three hours, that patient list is outdated. New data has come in. John gets to go home now because we actually have a placement for him. Well, how does, can you raise that to the top of your list? So the hospitalist doesn't come see them six hours later and you're like this now and there's so much potential and that's, that's fairly safe.
You're just gathering data. You're not, as you were saying, the more scarier things about, making a recommendation those kinds of things, which is, You know, a really exciting possibility as well. Just some of the administrative things that physicians and nurses do every day that's got me excited.
I don't think that necessarily has to be very far away either, using that technology. Or to be able to say, John wants an after visit summary. He, he's English speaking, but he's got a family member that's Spanish or Chinese, fifth grade Chinese level after visit summary. Boom, Be able to do that.
That's a great one, Brett. I love that one. I do want to say, I do keep saying scary because I don't want to come off as rah rah cheerleader for this because, like there are concerns that we will have to address along the way. The flip side of that is, I am not persuaded by a lot of the, the worried folks, let's put it that way.
When they talk about, things like just ambient documentation, just in general, well, how do we know that like the AI is accurate and what if it gets in there? And, we've got this cycle where we've got AI on the front end that is creating the note. And then we've got AI on the backend that's doing NLP and reading it for real world evidence.
And, we're creating this loop where there's, there's no human in there. And. My pushback on that is like, that is a valid criticism. At the same time, I have looked at my own medical record and in many cases, the stuff that's in there is. duplicative cut and paste. For me, most of the time it has been accurate, but I'm not a particularly complex patient.
I have the good fortune to be relatively healthy, so that's not an issue. But you know, I think it was one of the first studies of, of like what happens when patients have open notes. And I think it was 75% of them found at least one error. And of those, 25% of them were serious errors in the record. Like let's not kid ourselves that the manual processes we have right now are a foolproof way to have a safe and complete record.
Like let's just be real about what the problem is
and what our baseline is, right? Like the baseline isn't awesome always. I mean, there are some really good documents out there, but to think that we're going from, Perfect to now we're going to trust AI is not, I mean, I've used some ambient listening before in the clinic for several months and it's incredibly helpful.
There are things that get said, I'm a primary care physician in a primary care office that are just an aside. And that I would hope most of the time I get documented, but sometimes I don't because it wasn't really as relevant. Oh, I had a tetanus booster. I was in urgent care a couple of weeks ago, and I'll be reading some of these ambient notes that get created.
I'm like, oh, that's right. I think I would have forgotten that, so I mean, the argument can go both ways, but we you're right in medicine. We don't have the luxury of. Having a 50% error rate and thinking that's okay. Let's just work towards, you know, a lower, right.
📍 gosh, I really love this show. I love hearing what workers and leaders on the front lines are doing, and we wanna thank our hosts who continue to support the community by developing this great content. If you wanna support This Week Health, the best way to do that is to let someone else know about our channels. Let them know you're listening to it and you are getting value. We have two channels This Week Health Conference and This Week Health Newsroom. You can check them out today. You can find them wherever you listen to podcasts. You can find 'em on our website this weekhealth.com, and you can subscribe there as well. We also wanna thank our show partners, MEDITECH and Transcarent, for investing in our mission to develop the next generation of health leaders. Thanks for listening. That's all for now.