Intro: The following SHSMD Podcast is a production of DoctorPodcasting.com.

Bill Klaproth (Host): On this edition of the SHSMD Podcast, we talk about how to get beyond calling it AI, people, and explore how to use smart technologies to innovate. Isn't that what it's all about? So, joining me is Roberta Levy Schwartz. Roberta is a SHSMD Connections 2024 Keynote Speaker and FutureScan 2025 Subject Matter Expert, and you're going to want to come check out her discussion at SHSMD Connections 2024 in Denver, October 13th through the 15th. It's going to be fascinating. It's going to be informative. You're going to bring a lot of ideas back to work after this one. So, let's get into how to innovate with AI with Roberta right now.

This is the SHSMD podcast, Rapid Insights for healthcare strategy professionals in planning, business development, marketing, communications, and public relations. I'm your host, Bill Klaproth. In this episode, we talk with Roberta Levy Schwartz, Executive Vice President and Chief Innovation Officer at Houston Methodist Hospital. I love that title, Chief z Innovation Officer. Maybe someday when I grow up, I can be that. Anyway, sorry, Roberta, welcome to the SHSMD Podcast.

Roberta Levy Schwartz: Thank you so much for having me. I'm excited to be here and excited to speak at the conference.

Host: Yes. We are excited to hear from you as well. I know it's going to be a great keynote for us that you're going to put together. So, I'm interested in getting a quick preview from you for our listeners. So, let me ask you this, you believe that healthcare can be transformed through these new technologies. Of course, I'm talking about AI, patient-focused systems, and telemedicine, all of these things to help create the smart hospital of the future. I love it. So, tell us, why do you feel this way?

Roberta Levy Schwartz: When you look at the transformations that have happened in other industries, take it shopping with Amazon, no doubt, as well as other stores. My husband hasn't been to a store in probably eight years. When was the last time you hailed a taxi rather than calling Uber or Lyft? When was the last time you called a restaurant to make a reservation rather than booking something through OpenTable? Each of the industries have transformed. And whether it's marketing and outreach, identifying the kinds of things that you like and matching it up and saying, "Hey, you should look at this item if you looked at that item." They all use some level of intelligence to gather data around you. And there's probably nowhere where we have more information about people now, thank you to a lot of pushing that happened during President Obama in an electronic form. And so, the question is, how do we use this data? How do we use this information? How do we use telemedicine and other tools that are now at our disposal to transform the industry at hand, healthcare, our industry, and look at things differently, kind of turning that kaleidoscope. It has the same little shards inside. But when you turn the kaleidoscope one direction or another, you know, a 10 or 20-degree angle, it looks completely different. And I think that's what in our industry we need to do.

Host: I think that is so well said. And when you give those examples of, you know, "My husband hasn't been in a store in years", most of the shopping I do is on Amazon. And you're right, when was the last time you called a cab? It's Uber. So when you say those examples, you're like, "Oh yeah, these other industries have innovated. Maybe us in the healthcare world, we are a little bit behind." So, that really makes sense, Roberta. So, thank you for sharing that. And I know that you don't like the term AI. You say it's problematic. Why so?

Roberta Levy Schwartz: You know, I think the issue that you have is if you say to five people in a room, "Hey, what's an example of AI in healthcare?" Five people will give you five different examples from five completely different modalities. And then, they all talk past each other. So when you look at its basic definition, it's the ability to perform tasks where a computer can do those tasks quicker than possibly human intelligence or our visual or our speech or our decision-making can kind of come from. But when you then start to have a conversation and say, "Hey, you know, I think AI in healthcare, we should talk about AI in healthcare," it becomes so genericized that I actually lose myself in conversations trying to figure out what people are talking about.

And so, lately, we've had a lot of conversations about AI and the hottest topic that everyone wants to discuss is AI that's happening between conversations, doctor talking to patients, and the conversation is summarized and put in the electronic medical record. Well, I can have a conversation with one company that just does a summary of the record and gives you a SOAP progress note. I can have a different conversation with another company who is then taking that one step further and putting that information into discrete fields in the electronic medical record. I can have a conversation with a third company, and it's actually without even having the conversation, producing the progress note out of the information that's already available in the chart and with test records that have happened from that day. And if they all just use the term AI, I don't know which one we're talking about. And so, I force people all the time to say, "Okay, don't talk to me in the generic term. Don't talk to me in that very high-level words. Tell me what you're talking about, and then we can have a really great dialogue, which I will love."

Host: One hundred percent. So, let me ask you this. So, you kind of gave us the definition of AI. And you said, if you went to a party and asked five different people, you'd get five different examples or definition of what AI is. So, let me ask you, Roberta Levy Schwartz, what is AI to you?

Roberta Levy Schwartz: AI, it's anything that can indeed make the lives of our patients, our practitioners easier. Take the burden of simple human judgment or human tasks and take the burden onto the machine for doing the tasks that maybe humans either menially shouldn't be doing, don't have to do, or provide them with a level of information that makes their jobs easier. Now again, when we then go into it, I love having a discussion about AI in reading imaging, AI in producing notes, AI in looking at backend functions like access, like billing, AI in transcription, in translation. But in each one of these, we're going to be having a very different discussion about where technology is, how far have companies and technologies driven things, and where do I think there is noise in the data that causes it not to be as fully trustworthy as humans want it to be.

So in taking a super simple example, which I use a lot, which is the calculation of body mass index, like a BMI, right? Immediately, your computer is going to be able to do that calculation for you. Some call that AI, some call it just a calculation. I'm okay either way, right? But it takes the burden off a person who used to, way in the past, actually calculate that number. Now, it's just produced in the computer for you. What was complicated years ago to making that calculation now is we're up to the point where having the conversation and having the conversation between doctor and patient and producing a note works really, really well for our medicine physicians and our primary care physicians, but is really more complicated for our surgeons where templates work much better than conversational collection of information into a record. So, we then have to get down and peel that onion back until we get to the point at which both we're having conversations, similar conversations, but we can also talk about within each one, which is good and which is bad, and which still has opportunities and which is now mature. And we can have high-level conversations about things like bias. But honestly, having a conversation of bias in data that we have about a vital sign collection is going to be different than bias that we're going to have when it comes to an AI bot that is doing screening for cancer, or doing screening for other types of diseases. So, we need to talk in very specifics about each one.

Host: So, Roberta, what I'm getting from you is we need to have a discussion on how AI can be used in that specific situation. You keep talking about, you know, when you have this discussion, we need to have a conversation. So for you, I'm just wondering how have you implemented this at Houston Methodist Hospital, do people come to you and say, "Hey, Roberta, I want to use AI for this. Can you help me out?" And you say, "Okay. Well, let's talk about this." Or are you going to people saying, "Hey, I've got this idea for AI. Let me talk to you about this." Does that happen at Houston Methodist?

Roberta Levy Schwartz: Yeah. So, we have a wonderful model where we have recognized that the industry and startup companies and big companies will always move faster than our development. And you know, I love doing some development in partnerships with academic partners like Rice or University of Houston. It's really incredible. But that's going to be filling in the pieces where other companies are not particularly interested. So when we found companies who are moving very quickly in a space of twinning our data against the landscape of 300 billion Medicare records and looking at big data there, or BioButtons by companies like BioIntelliSense that stick them on patients and actually collect trended vital signs to look at whether we can diagnose issues earlier or better, or work that we're doing with companies like HDAI in the space of looking at our data for opportunities after patients leave. We've worked with these companies and many, many, many others to find out whether or not there are opportunities to improve care. So, we take the work that they're doing and we put it through its paces and we see whether or not care is better, burden on clinicians is lower, the information that we're getting is better, we need, in many cases, less people to do the work that we were doing before. So, each one of the technologies we measure for outcomes. So, we are so impressed with the startup community and now maturing community to look at how we can provide better care, less burnout, and be able to prove that out in large-scale models. So, anything we do here, we're going to do across eight hospitals and across a significant number of inpatients or outpatients to be able to test whether or not these technologies have strong outcomes.

Host: So, do you find yourself leaning into AI more and more, looking at problems and the potential solution is AI?

Roberta Levy Schwartz: I find myself oftentimes looking at problems and saying, "Are there companies who have identified unique ways of solving the problems I have without just adding people?" Right? So, I'll give you an example. When we rolled out the vaccine program here at Houston Methodist, we were one of the largest providers of vaccines in the state of Texas. And we, in a very short period of time, delivered over a million vaccines. We were one of the sites the state identified that were going to distribute the vaccines. And we knew that people in the beginning were going to be phone calling anybody to get an appointment. So, we pictured a point where our phone systems were going to be basically taken down, just extreme impacted by the amount of people who are going to call for vaccines. And we teamed up with basically a phone bot provider who could take the patient's intake, qualify them for whether or not they qualified for the vaccine, depending on age, depending on where they were in kind of the pecking order that the state had defined for distribution and then setting them up with an appointment. And we did that all without human intervention, thereby avoiding our entire phone system crashing while trying to take in a million new appointments in a very short period of time. And we couldn't have done that by ourselves, right? We've worked on it with a partner.

Host: Yeah, that makes sense. In your keynote, are you going to be giving other examples like this? Because I think examples like you just shared with us are very important and helpful.

Roberta Levy Schwartz: Absolutely. You know, I believe that we have to talk about not the theoretical, which is oftentimes where we get tripped up, but the reality. And I think one of the things the government has said, is take reality and make it better. So if you identify issues and problems and biases within the models as you're building them, fix them. Commit to fixing them, which is absolutely right. But we're not going to be able to get it perfect on a theoretical level. We're only going to get it better when we really start putting it in, pushing these technologies to their limit, and then asking them to do more and asking them to do it better, right? That's our job.

Host: Yeah. So, let me ask you this then, switching gears just a little bit. So, this podcast generally is a healthcare marketing podcast. We generally delve into marketing subjects, but time to time we veer off of that. So, let me ask you this for this discussion you're going to have for us. I know you're going to share more about digital innovations that have proven helpful in marketing, communication, business development, and strategic planning. Those are the things we cover on this podcast. Can you give us an example or an insight into a few of the things you're going to share about marketing and new technologies?

Roberta Levy Schwartz: Well, I think the beauty of marketing is that when you can find the right population with the right needs and market to them to meet those needs, we are all better off. So, the question really comes down to, you don't want me to give you a list of a hundred thousand patients if you're marketing to Crohn's and colitis, right? You want me to find the Crohn's and colitis patients, right? You want me to find that. You want me not only to find the Crohn's and colitis patients, but if I'm doing a really good job, you want the Crohn's and colitis patients who have the next need, who have missed their appointment, and be able to segregate those people at the moment in time where they need care and be able to serve up that information about the care.

And that's a lot of what you're talking about here. You're talking about taking that information and finding-- I'll give an example from the work we've done with HDAI. And that is our physicians basically, our surgeons, they kind of have set of numbers that they give to everybody. I will follow up with you in seven days. I will follow up with you in 14 days. I will follow up with you in 30 days, right? I mean, they just provide the stock number. And what we found is that I maybe needed to follow up in 24 hours and you could have waited 18 days, right? It wasn't one-size-fits-all. And as we segregated that data, we were able to find eight patients who needed to be seen within 48 hours. So, we didn't need every patient to be seen in 48 hours. We needed eight of the 100 patients to be seen in 48 hours. And we could focus on those eight patients and getting them in sooner.

So, from a marketing perspective, the more I can feed that type of specificity to a marketing team, for them, it's like gold. They now know who they're targeting, what they're targeting them for. Not everybody needs a lung scan. No, no. These people have been recommended to get a lung scan in the next, you know, six months. Focus on these people and getting them the information they need. These are the high-risk people. Focus on my risk. You are at high risk, serve up this information.

Host: Yeah. That's very well said, and I think you hit the nail on the head. And that's why this is going to be a dynamite discussion with you at SHSMD Connections 2024. And anytime we get a chance to have a Chief Innovation Officer with us, you know you're in for a great educational session. There's going to be a lot of wonderful takeaways. Before we wrap up, Roberta, I want to thank you so much for your time. Anything else you want to add?

Roberta Levy Schwartz: I think I love when people challenge things or ask about new examples or come up and say, "Here's what we've done." I was reading this morning about the work at Johns Hopkins in their emergency room and how they've used AI to better triage patients. I think the more we all talk about what's out there and the opportunities before us and raise their hands and say, "Guess what we've done and guess what we're seeing. And it's not perfect, but we think there's an opportunity here," we will raise the entire field. You can argue whatever you want to about the changes that happened in the taxi industry. But I can tell you that I think we've raised all boats to be more available, more accessible, easier to find out where we're going and how long it's going to take us to get there. You know, we need to help each other. It's only with helping each other. It's not this field right now is not exceedingly competitive in the hospitals accepting this, I know it's very competitive in a lot of the startup markets, but because we're all just so hungry to do it better. And I think all of us talking about it makes us better.

Host: And that is the goal. Roberta, this has been fascinating. I know this is going to be a great keynote at SHSMD Connections 2024. We really appreciate your time today. Thank you so much.

Roberta Levy Schwartz: Thank you so much for having me.

Host: And once again, that's Roberta Levy Schwartz. Roberta is a keynote speaker for this year's SHSMD Connections 2024 in Denver, Colorado, October 13th through the 15th. Stay tuned, people, for early bird registration. And if you found this podcast helpful-- And how could you not? Come on-- please share it on your social channels. And please hit the subscribe or follow button to get every episode. Sorry, been a long day. And to access our full podcast library or other topics of interest to you, visit shsmd.org/podcast. This has been a production of Doctor podcasting. I'm Bill Klaproth. See ya!