AI's Role in Redefining Medical Professional Development with Dr. David Rodgers

Unlock the future of healthcare education with us at the Sim Cafe, where the fusion of artificial intelligence and medical simulation is not just imagined, it's happening. Our special guest, Dr. David Rodgers from Indiana University Bloomington, brings the awe-inspiring prowess of AI to light and discusses its monumental impact on interprofessional learning. Journey through the corridors of Indiana University's innovative health professions campus and discover how accreditation mandates are ushering in a new era of collaborative learning that's reshaping pre-licensure education and professional development.

Get ready to challenge everything you thought you knew about medical scenario design as Dr. Rodgers breaks down a ground-breaking study from Simulation in Healthcare. With ChatGPT at the creative helm, the study reveals a blend of promise and caution - scenarios burst forth with potential, yet peppered with imperfections that only the human touch can refine. We entangle ourselves in the fine threads of human-AI collaboration, marveling at the intricate dance that ensures medical guidelines are not just met but embodied in simulation design. It's an intricate pas de deux of man and machine, where the artistry of human expertise guides the algorithmic precision of AI.

Witness AI reshape the landscape of medical education, stretching its digital tendrils into scenario development workshops and beyond. Dr. Rodgers recounts firsthand experiences from IMSH, where the speed of AI left attendees both spellbound and contemplative of the technology's capabilities and limitations. As we contemplate the horizon of healthcare, we delve into AI's burgeoning role in medical imaging and its necessity in training tomorrow's professionals. Grasp a deeper understanding of how to navigate the ethical tightrope of AI in education and join the conversation on its responsible use in an ever-evolving academic world.

Innovative SimSolutions.
Your turnkey solution provider for medical simulation programs, sim centers & faculty design.

Transcript

Disclaimer/ BrainNet Ad/ Intro: 0:00

The views and opinions expressed in this program are those of the speakers and do not necessarily reflect the opinions or positions of anyone at Innovative Sim Solutions or our sponsors.

Deb Tauber: 0:13

Are you embarking on a medical simulation center project, whether it's from the ground up or enhancing what you have? Brainnet Consulting is your trusted partner. Their expertise isn't just in design. They ensure your center has the necessary infrastructure, optimal layout, cutting-edge technology and is primed for accreditation. With 29 successful global projects, they know what it takes to create a center that's ready for tomorrow's education needs. Contact BrainNet Consulting. We're building the future of medical simulation today.

Disclaimer/ BrainNet Ad/ Intro: 0:51

Welcome to The Sim Cafe, a podcast produced by the team at Innovative Sim Solutions, edited by Shelly Houser. Join our host Deb Tauber and co-host Jerrod Jeffries as they sit down with subject matter experts from across the globe to reimagine clinical education and the use of simulation. So pour yourself a cup of relaxation, sit back, tune in and learn something new from The Sim Cafe.

Deb Tauber: 1:31

Welcome to another episode of The Sim Cafe, and today we are so fortunate we have Dr Rogers. He asked that we call him David Dave. So David Rogers is the Director of Interprofessional Simulation Center on the Regional Academic Health Center campus of Indiana University, bloomington. He's also a clinical assistant professor of medicine in the IU School of Medicine. He's actually a paramedic and he served time as a flight,EMS paramedic, supervisor, manager of hospital-based critical care transport services and he has over 25 years of simulation-based education and he is going to enlighten us today. So welcome Dave and welcome Jerrod. Thank you for being on, and good morning, good afternoon or good evening.

Jerrod Jeffries: 2:24

Yes, yeah, and thanks for being here with us, dave. I'm excited to dig in. I know that there's a lot of buzz around AI and I think that I've heard your name come up quite a few times when we hear about artificial intelligence or AI. Of course, that could mean generative AI or just strictly artificial intelligence, but I know that we want to get into quite a few things here, so thanks for being on. And first, I mean, what are you working on today? 20 years is a long time, or maybe you want to start with even a little more background if you want to fill us in on anything before we jump into some of the questions.

Dr. Rodgers: 2:57

Well, as Deb mentioned, I've been doing simulation for a long time and was actually an original first year member of the Society for Simulation Healthcare, so when it started I was one of the first year members with that. So in that time a lot of things in simulation have changed, and the introduction of artificial intelligence in a couple of different flavors has the potential to change it even more. Flavors has the potential to change it even more, particularly in regards for developing simulations as well as how we use data that's refined by artificial intelligence.

Deb Tauber: 3:34

Why don't you share a little bit more about your background to our listeners, just where you're at right now, what you're doing?

Dr. Rodgers: 3:41

Yeah, so right now, as you mentioned, I'm in Indiana University, Bloomington, which was a great opportunity. Here they re-envisioned health professions education as they built a new hospital here and a new medical school, school of nursing, school of social work and speech, language and hearing, because all of those are combined on the same campus. So it really was a really attractive opportunity to have a brand new simulation center with a new focus on doing interprofessional education, and that's kind of what drew me here. Usually, if you look at a health professions university, the school of medicine is on one side of campus, the school of nursing is on another side. Who knows where all the other support schools are? And it's very difficult to do in a professional education just because of logistics. Here it's just so much easier to do because everybody's in the same building, it's everybody's home and when you go to the simulation center you're not going into somebody else's territory, such as you're going to the School of Nursing Simulation Center or the School of Medicine Simulation Center. You're going to your simulation center.

Jerrod Jeffries: 4:44

So for some of our listeners who might not know, there's IU, indiana University in Bloomington and then IUPUI, which is in Indianapolis Indiana University, Purdue University in Indianapolis and they have a unique model with Fairbanks as well. Yes, is this somewhat similar to Fairbanks in Indianapolis or is it completely separate model i?

Dr. Rodgers: 5:03

It is. I like to tell people we're exactly the same but completely different. So the way that both of our centers are funded is through a combination of school medicine, school nursing and health system funding, but they all pay for a percentage of the budget based on their utilization. We're exactly the same that way. What we're different is the Fairbank Center's actually based inside IU Health and, because they have to sit somewhere, I'm based here in Bloomington we're based inside Indiana University, and so our paychecks look different. They come from different organizations, but the money that goes into our paychecks and goes into all of our operating expenses is coming from the same sources.

Jerrod Jeffries: 5:48

And it's two Indiana schools now that I've heard are doing that really lean into the interprofessional education, and I'm sure this is done elsewhere I'm probably just not knowledgeable of it, but I think it is much more impactful for the student, for the participant, in my eyes, yeah, and when you look at the role of interprofessional education, both in pre-licensure education and professional education, there's more of an emphasis on interprofessional education.

Dr. Rodgers: 6:17

There are elements of the school of medicine for any medical school's accreditation, as well as any nursing school's accreditation, where you have to have some level of interprofessional activity. And then even when you get into post-licensure education, particularly in residency programs, almost all the residency programs require some degree of interprofessional education. So there's mandates from the accrediting agencies that are really pushing a lot of this, as well as the fact that people realize it finally makes sense.

Jerrod Jeffries: 6:48

Yeah, certainly, I mean certainly, it's about time yeah.

Dr. Rodgers: 6:53

Yeah, yeah, the old model of of training all health professions was you know, you, everybody went to school together in high school and then, split out and you got your professional health professions, education, and then at the end you kind of merged all together again in the workplace. And that was just not a good model, because nobody works alone in the workplace and so getting the Especially in health care.

Jerrod Jeffries: 7:19

Right, especially in health care. This is true for everywhere, but I mean health care even more so. So is your work, then, dave. Is it being spread across all these different interprofessional areas, or are you more targeted within school medicine, or more leaning towards nursing, or how does that look?

Dr. Rodgers: 7:36

Yeah, all our staff, including me, are stretched out across the board with all of our users.

Dr. Rodgers: 7:41

I have a faculty position in the school of medicine and I have some responsibilities as being a faculty member, but I spend just as much time working with the school of nursing and with the health system as well, as well as some other smaller partners such as the school of social work, speech, language and hearing.

Dr. Rodgers: 7:58

And then you mentioned Indiana, particularly between Indiana University and Indiana University Health, which are two separate entities and they do not cross. There's no matching point at the top, the scope of simulation that's done throughout the state because they all have multiple campuses. With medicine nursing, there's nine medical school campuses, there's nine nursing school campuses. There are I don't know how many IU health facilities 15 or 16 major facilities and the scope of simulation with that is just amazing. In Indianapolis alone there's close to 10 independent simulation programs that support either Indiana University or IU Health and they each have their own niche. You mentioned Fairbanks, which is the big one, and it covers a lot of territory for all three of their major players medicine, nursing and the health system. But there are some other specialty areas that are more focused on doing some local activity or some specialty-specific activity.

Deb Tauber: 8:57

What do you see? I know you're on the cutting edge with AI in the center, and why don't you tell us a little bit about what you see is the potential for AI and what you guys are doing with it there?

Dr. Rodgers: 9:09

You know AI has been around a long time and what really changed in the past year and a half has been that it has become accessible for everybody. Prior to that, if you were using AI and you were using something like IBM Watson or something like that, you were a programmer. You were heavily invested in some technology. With the creation of primarily, chatgpt, that's been the one that's a driver. There are other models out there that are similar. You know Microsoft makes one now, so there's just there's a variety of them. Adobe has one, but with ChatGPT one because it was free and everybody could access it and you didn't need some sort of special skill to learn how to use it. All you had to do was know how to type and ask a question and then from there the AI just did its thing. And so that really changed a lot of how we approached AI within simulation.

Dr. Rodgers: 10:07

Here in our department, one of our part-time techs who's a full-time tech up in Indianapolis at the Fairbank Center, Alex Robinson he found it shortly after it was introduced started showing us some things it could do. We started looking around with the rest of our staff and said you know that we could do some cool things with this and particularly in the area of simulation, scenario development, and find ways to use it to do that. And so we did and we published a study that was in, I think, the December issue of Simulation in Healthcare this past December, where we did a study on using ChatGPT specifically as an AI entity to build a scenario, and we built two scenarios. We had a group of people that we blinded to what the origin of the scenario was. We put it all into a standardized template and had them grade it basically and they said, yeah, this is pretty good. This is like B minus work.

Dr. Rodgers: 11:01

It wasn't outstanding, there were some things going on, but it looked pretty good. And then we revealed to them that a human didn't write it and it was like, oh okay, that's interesting. And so we got feedback on that. We all contributed into the article and it did some things just exceptionally well and just amazing. How much time do you put in as a simulation designer in just developing the backstory? I don't know how many minutes I've spent just on a single simulation. What's this guy's name, or something like that? And we just would ask it a question. It created a name, it created a whole backstory for them and was just amazing and it did really well on creating objectives based on what our learning goal was. So it really did an exceptional job and I would say about 90% of what it produced you could just take as was and just plug it in.

Jerrod Jeffries: 12:01

Wow, that's high.

Dr. Rodgers: 12:02

Yeah and the, but the 10% that it didn't do well in it really did a horrible job. And so one of our simulation coordinator, Matt Needler, when we were working with the program initially, he says, wow, I'm out of a job, you know. We looked at it and said this is, you know, this is doing a lot of the stuff I do, this is doing a lot of the stuff I do. But then as we got to use it more, we realized it was just another tool that we needed to use to make our job more efficient and more accurate. It did not replace us. I don't think that's going to happen for a while until we build up some more definitive use and how it addresses accuracy, because some things it was horribly inaccurate on and it dropped these things in that just didn't make sense and you needed somebody to edit that and realize what the limitations were. It was. There's a couple of different words for it. One is hallucinations, one is confabulations. That happened, where it just drops in a sentence that makes no sense, such as one of the sentences we had was a V-fib cardiac arrest and it put in that there was a V-fib patient, was in V-fib, showing flatline on the monitor. Okay, that doesn't work. But so we figured chat GPT. Watched too many medical TV dramas on television where they shocked us instantly, so that's all we could figure, but there were inaccuracies like that that dropped into it.

Dr. Rodgers: 13:32

One of the techniques that you really have to develop now to use chat, gpt or any of the AIs in that realm is learning how to write prompt questions. That's a new skill set and it's something we've practiced. We've gotten better at it. We realized that we made things. You need to be specific For one of our scenarios. We asked, we told it we want this scenario to follow American Heart Association guidelines for cardiac arrest management. It did. Problem was they weren't the current guidelines and so it gave us inaccurate doses. It gave us 20 doses instead of the updated 2020, at the time, 2023, there was an update, I think it went back the whole way to 2015. It gave us 2015 guidelines which actually answered our prompt. It used AHA guidelines. So we realized we had to be more specific in the prompts and work with that. Also, we ran two different scenarios One was an adult scenario and one was a peds scenario when we did the study and the results of that kind of mimic.

Dr. Rodgers: 14:34

What's going on in some other areas in that chat. Gpt in medicine does pretty good at broader scope, general topics such as adult cardiac arrest. There's a lot more out there about adult cardiac arrest than there is maybe about a pediatric case with pediatric asthma management, severe asthma case and the asthma case scored lower than the adult cardiac arrest case. This would be consistent because with what we're seeing in other places in the literature that if you ask it a very broad question that applies to a general medical topic, it does good. But if you start narrowing it down to a very specific thing and one of the articles was on ophthalmology it did pretty good at general questions. Then when it started getting in some more disease specific it just was losing it. It couldn't feed back accurate information. So again, it's going to be a tool to help build simulations. It'll be a good starting point but it's not going to replace anybody. No one's going to get to have a job quite yet Not yet.

Deb Tauber: 15:35

Not yet, not yet Now. I just want to plug a little. I know you did a webinar for Lance, so if any of our listeners are interested in learning a little bit more, you can go to healthysimulation. com and find Dr Rogers' talk on AI. It was very well attended and it was a great talk.

Dr. Rodgers: 15:52

Oh, thank you. One of the things we mentioned in that one and going back to the accuracy is that I know it's a computer program, it's not human, it doesn't have human traits, but it wants to give you something, even if it doesn't know the answer. It'd be like somebody making yes, you ask a kid in elementary school don't know the answer, so I'll make something up and ChatGPT will do that. It did a horrible job when we asked this to give us 10 references to support this case in the debriefing. So we had 20 references, total 10 for each case.

Dr. Rodgers: 16:30

Out of the 20, 16 were completely fabricated. They did not exist and they were incredible fabrications because I looked at them first and they looked legit. I said I know these people. They put people's names in the citation that I knew it was on topics that they write about and it was in journals that they would publish in what. And it even got better because there were a couple who were made up for simulation of healthcare and so when I went and cross-referenced them and looked them up, it would say and I can't remember the exact numbers, but it was in volume 12, issue six, page numbers this. And when I went to volume 12, issue six. Those page numbers were included in that volume, but the article wasn't there. It was actually overlapping a couple of other articles and so it actually refined it to the point to giving page numbers that were accurate to the volume and issue for that journal, which is stunning. So it knew what it was doing but it basically just lied it made it up.

Jerrod Jeffries: 17:37

I was just waiting for you to be like and I was one of the references that I've never written that.

Dr. Rodgers: 17:42

So this year's IMSH in January in San Diego I was in a meeting. We had a breakout meeting it was from one of the groups I'm in and pulled the article out and one of the made up authors was there and it was Dan Raymer and so I asked Dan. I said, hey, dan, do you remember this citation, this article? And he's looking at it. And then Jeff Cooper, who works with him quite considerably, comes over and looks over his shoulder and Jeff says Dan, I didn't know you had an article published in Chess. Dan says I didn't either, and it was just completely made up, but his name was on there, it was on a topic that he would talk about, and so it was just amazing, the things that came back.

Deb Tauber: 18:26

So I think that gets into some of the limits of AI. What other limits do you see AI is having?

Dr. Rodgers: 18:33

Yeah, right now the biggest one is accuracy being able to be accurate in giving us factual information. Some of that's going to come from the prompts itself, how you ask those. Hence our issue with getting outdated accurate information when we asked about ACLS guidelines from American Heart Association. So you've got to be on the ball for that and paying attention to it. But the accuracy is, hands down, the biggest issue.

Dr. Rodgers: 19:01

When we did a workshop at IMSH this year on it and we had several people in there Some of them were using their phones, others tablets, others computers and we had them actually develop scenarios and they were just shocked that in a space of two or three minutes and a couple of questions, they had a fully fleshed out scenario that they just needed to go through and edit. They had the situation, they had the goals and learning objectives, they had an equipment list, they had how the case flow all that and it just needed to be double checked. But essentially it produced something that would take a typical SimTech writing a scenario or SimEducator a couple hours. It did it five minutes and so it really streamlined the process. So, looking at accuracy, the other thing it's going to have difficulty with is that it doesn't think like a simulationist. So when you typically sit down and write up a scenario, it's typically a branching scenario that looks like a flow chart.

Dr. Rodgers: 20:01

If they do this, it goes here. If they don't do this, it goes here. None of the AI programs can do that. It's going to give you a text-based progression, and so trying to translate that into the various states that we use or frames or whatever software you're using calls it Translated into those frames. It's just not there. So that's going to be another thing that the simulationist has to do. They've got to take what it gets. They get a good backstory, they get a good front end, they got to outline a patient progression, but it doesn't match the way we think, particularly if we're going to program it.

Jerrod Jeffries: 20:36

A question I have around all this, which I think limitations, as it is today, right, even in six months, who knows, or two years, it's going to be a completely different ballgame. It moves that quick. But my question is around the students. So are students or participants becoming more enthralled or involved in simulation due to this technology, or is it? I only want to be a nurse, I only want to be this type of doctor, I only want to be, and so they kind of push away for the technology. Are they leaning in more that you're seeing now?

Dr. Rodgers: 21:04

When you look at AI in a broader scope and you look at AI in healthcare, any of the people in pre-licensure training now have got to learn what it is, what it can do, what it can't do. I was at a meeting, as Deb mentioned, last couple of days. I was in Chicago and we were at a meeting and the person sitting next to me said you know, he had went to a presentation by one person on AI and medicine and he said you're not going to be replaced. He was talking to a group of physicians. You're not going to be replaced by AI. What you're going to be replaced by is by a physician who knows how to use AI. So that distinction was there. That again, it's a tool and you have to learn how to use it. And so I think in all the professions you're going to see AI in a variety of different ways coming up looking at it.

Dr. Rodgers: 21:52

We're already using AI in medical imaging extensively, where it can scan x-rays or other medical images and identify abnormalities based on pattern recognition. They've got programs out there now that well, they've been around a long time. I mean, if you hook up and do a 12 lead with using, like a Zoll or a LifePak, the Fibrillator in a pre-hospital world. It prints out the 12 lead. It tells you what it thinks it is. There's not a little guy inside there that's reading it, it's AI that's reading it and looking for pattern recognition to tell that. So we're already using AI in a broader sense and as the power of AI develops, it's just going to become more integral to what we do as health professionals.

Deb Tauber: 22:38

Now, what about as far as an educator, as far as a teacher goes? What do you feel about learners using it and what are some of the things that you guys are trying to navigate around? I'm sure it's got to be a hot topic at the school.

Dr. Rodgers: 22:52

Yeah, within education and this covers not only higher education but elementary, secondary education is how to use AI appropriately, and they're actually running college courses now for students on how to use AI appropriately. Again, it's another tool. If I'm a student, I want to use as many tools as I can. Is it cheating that I use Google to find a reference Because I didn't go to the library and go to the card catalog? I don't even know if those exist anymore, but you know. So, as you progressed with the technology in education, it just has to go with it, and there are always limitations. I mean, they've already got the plagiarism checkers out there so that people can tell. If you copy and paste it somewhere, Google and put it into your term paper, that's going to get caught. So but Google or other search engines are still a source that an individual can use, a student can use to help build their project. So they just have to know how to use it responsibly and ethically, and I see AI being that as well.

Dr. Rodgers: 23:55

Ai has some good opportunities for doing things outside simulation and education. One of the things is that you can ask it to explain something to you and so maybe what you have as a source or what your professor talked about. Your teacher talked about didn't make sense. You can ask it the same question and it'll spit something back and give you information on it. So, OK, that makes more sense. Or you can ask questions. You can query it to go out for more details. Again, there's a issue potential there of accuracy, so you can't take it on face value, but it can help inform some of your decisions into building products later.

Deb Tauber: 24:32

Yeah, I totally agree. I think it's helpful and I can see where an instructor or teacher has just so much knowledge where they're trying to break down a subject to a learner and they might be over their head, whereas if the learner just types something in, they might get an answer that makes a lot more sense to them.

Dr. Rodgers: 24:49

Yeah, and you'll see one of the places that you'll see AI used a lot and particularly in secondary education, higher education is that you can create something, but you know how. You're going into a typically online and you write something and it says you're limited to a thousand characters and you've got 1120. You could take what you wrote, put it into an AI program like ChatGPT and say, narrow this down. So it says the same thing but doesn't know a thousand characters, and it'll do it, and so it helps you edit your work and figure out how you can condense something and not lose your point. So there's applications for that. It's still your work, you're just having it help you edit it and makes it more time efficient.

Deb Tauber: 25:34

Excellent, excellent.

Jerrod Jeffries: 25:35

And I also love your analogy with the library. Do decimal system or card

Dr. Rodgers: 25:40

catalogundefined.

Jerrod Jeffries: 25:41

Cause it is funny. It's like I mean, there was a point I don't know what year that would have been, but where it's like you have to go to the library to find this, this article or this or this research. You know what book did it come out of? And then search engines came along and it's like, oh, this is much more efficient. Same thing with this tool, right, it's just much more efficient and, to your point, it's not always accurate. That is a limitation, and it's just like you use it or you don't, and you embrace the future or you kind of get left behind with it no-transcript.

Dr. Rodgers: 26:32

The other place that you're seeing it again in healthcare there's a lot of uses for this is you create a document but you want to change the reading level. Maybe you wrote it at the graduate student reading level, but you need this to be understandable at the 10th grade level. You can throw your document in there and say make this at a 10th grade reading level and it changes your wording, takes out some jargon, it recrafts it and turns it into the same thing but for a different audience. And that's another use that you could see, particularly in patient education. You may have patient education materials that are a little too technical. You can put it into a program like CHATGPT and say make this for a person who is, for an adult who has not graduated from high school A lot of people like that and so that it'll change the wording around and eliminate some of the jargon, some of the six syllable words, things like that, and make it easier to read at a lower grade level.

Deb Tauber: 27:31

Can it translate?

Dr. Rodgers: 27:32

It does so you can translate. There are I mean, they've had programs do that. Wow, google Translate. I use it a lot. Yeah, when one of my Spanish language friends has a birthday, I go into Google Translate, I type my happy birthday wishes in English and translate that to Spanish and I copy and paste it into the text I send them. But so I cheat on that, but it does translate. It has chat, gpt I can't remember how many languages it's in right now, but it does do a lot.

Deb Tauber: 28:01

Now what's down the road? What do you see down the road for AI? Oh boy, Wide open. Oh boy, wide open. Well, and I think you've got such a great perspective, having been, you know, like you mentioned, at one of the first meetings, at IMSH, at the first meeting of IMSH, so you've seen the trajectory of things going over 25 years. So where do you see not that you have a crystal ball, but where do you see things going?

Dr. Rodgers: 28:28

So there's some logical ends to it right now that it's going to OpenAI. The company that makes ChatGPT is getting ready to come out with a new product called Sora. Sora develops videos based on text prompts and the quality of the videos is unbelievable. It looks like there's a Hollywood filmmaker that has recorded this and it gets reflections and shadows, everything. It's just an incredible video and we're looking forward to that product coming out to become available to us.

Dr. Rodgers: 29:01

At the same level as ChatGPT is because we can say you know, our mannequins don't look like real patients. There's a limit to what we can do with them, particularly if you're trying to show and I'll go back to the pediatric asthma case we did to show a 10-year-old with high work of breathing in the emergency department. We would be able to ask the OpenAI SOAR product give me a video of a 10-year-old child in a severe asthma attack in the emergency department, and it will pull together a completely AI AI generated video that we can play at the beginning of the scenario to set the stage, because the mannequin I don't care how much whose mannequin you got for a pediatric patient you can't make them look distressed. There's a limit. The work of breathing has a limit to it and so you can really do something like that. So AI generated video, I think has terrific opportunity in the context of simulation.

Deb Tauber: 29:59

Yeah.

Dr. Rodgers: 30:00

The next place. I know it's going there because I've talked to developers who are already working on this. So with some of the high technology simulators you have now, you get data out of it how good were your chest compressions, how long did it take to do the defibrillation, how long did it take to give medication X? All that can come out of the simulator software and help you in the debriefing. The next stage of that is to do AI analysis of the event itself with what was said in the event. So there's already people working on having recording the audio which with an AV system, you're recording anyway. Ai is sensitive enough that it can pick out individual voices and say person one, person two, person three, and come back around with that and then you can start developing it to actually worry and say, find a good example of closed loop communication, and it'll find that in the video or in the text and do that. It may get to the point of actually grading how well the team interaction was based on a clear leader, use of names, use of direct messages, closed loop communication call outs. It's going to be able to identify those things and at the end of the scenario, as you sit down with the learners. Right now you can get a score out of some of the simulators for how well you did CPR. We're going to be able to get a score on how well we did, an objective score on how well we did in managing the team with leadership and followership and teamwork, and it just be AI generated from that. So that is a tremendous potential.

Dr. Rodgers: 31:37

Taking that one step forward nobody's working on this one yet, but people are thinking about it is to look at the debriefing record, the debriefing, have it fed into an AI so the briefer can then query the AI and get feedback on how well they did as a briefer. Again, you can score it for some things so it identifies who the briefer is, just because it can identify who's in charge or who it thinks is in charge, which could be another problem, because if it thinks it's somebody else, then that briefer has an issue. Could be another problem because if it thinks that somebody else and that debriefer has an issue. But we'll start outlining and said I noticed you asked more. You said more statements than asked questions. That's an important point for a debriefer to know that they're talking too much and lecturing and looking at that, or it'll look about what type of questions do you ask? Too many closed questions versus open questions. You concentrated on learner number three and ignored the rest of the people. So we'll actually be able to get feedback from the debriefing that a debriefer could use all by themselves. Just go in and ask it. They don't need to necessarily have a peer debriefer there with them because that takes extra time and it pulls up faculty time. Some faculty don't like to give feedback to a peer, so this gives you a more objective state.

Dr. Rodgers: 32:54

And then, on the short-term future of it we're already seeing this there's a couple of products already in the market for helping build the backstory of the patient so that you can have a voice interaction. So it's basically voice to text, text to voice, and so that you can walk in and you can ask your simulator hi, what's your name? And it'll tell you his name, like your birth date, it'll tell you the history and it can get detailed as much. If you're like, if you're trying to start an IV on your mannequin and you're trying to distract him says, do you have any pets at home? So oh, yeah, I have a dog. What's his name? Fido? And you could actually put that information into the backstory and create a real-time conversation.

Dr. Rodgers: 33:34

One of the products that's on the market now actually has an evaluation system in it with their patient interaction. It will give you a point if you introduced yourself, so it knows. If you didn't introduce yourself, you don't get that point. Did you check two sources of identification? Did you ask the name? Did you ask a birth date, things like that? So it will actually analyze it and put it into a checklist format. When you're done, you may have 10 items on your checklist and you'll find out. You got an 80%. What did I miss? What did I forget to do? Because the AI can analyze that. So a lot of things we can do in quality improvement assessment, a lot of territory in that. That's going forward.

Deb Tauber: 34:16

Excellent, Excellent. Now we really appreciate your time today and all the things that your contributions to simulation, and do you have anything that you want to leave our listeners with? Any words of wisdom? Any final thoughts?

Dr. Rodgers: 34:24

I mentioned earlier. I just want to highlight this again it's a tool. It's not not an end. It's a means of us to get to that end by using all the other tools we have available as simulationists, which includes our brains, in our own common sense. So figuring out how to use that tool to our best advantage is is that's the skill set that comes up, and it's a it's a new skill set that the simulationist is going to have to attack.

Deb Tauber: 34:49

Yes, I would agree. I think that if you are thinking that AI is just going to go away, I don't think that is going to happen.

Dr. Rodgers: 34:58

No, it's here to stay and it's just going to become a more integral part of what we do, not only in simulation-based learning but in healthcare in general. It's going to have to be there, one of the kind of referencing, the ability for it to analyze conversations and talking with some physicians earlier this week. There's products in development now that will voice record a code and then at the end transcribe it for you, so you don't have to sit there and write it out. It knows the time. So if you gave, if you heard somebody in a closed-loop communication, epinephrine given at 7.38 am, it records that and you don't have to worry about it and track all that. And so, and particularly if it can combine with other elements, such as the feedback you get from some of the CPR feedback devices, what you're getting from the EKG monitor, and integrate that all into a cohesive package, that really requires very little effort. It's going to require somebody to sit down and proofread it.

Jerrod Jeffries: 36:00

Or a different alternative AI program.

Dr. Rodgers: 36:03

Yeah, so I think within all of healthcare and that's another thing we're going to have to keep up on is what is healthcare using AI for? Because we've got to make sure we're integrating that into our simulations so that the fidelity of our simulated environment is going to match the steadily increasing capabilities of the clinical environment.

Deb Tauber: 36:25

Yes, well said Well. With that, we really want to thank you and we will see you next time.

Dr. Rodgers: 36:31

Okay, thank you for inviting me. I enjoyed the conversation, thank you very much.

Disclaimer/ BrainNet Ad/ Intro: 36:43

Thanks for joining us here at The Sim Cafe. We hope you enjoyed. Visit us at www. innovativesimsolutions. com and be sure to hit that like and subscribe button so you never miss an episode. Innovative Sim Solutions is your one-stop shop for your simulation needs A turnkey solution.