EMTalk Episode 4: AI: Artificial Intelligence in EMS

EMTalk Episode 4: AI: Artificial Intelligence in EMS

EMT-Paramedic program

EMTalk Episode 4: AI: Artificial Intelligence in EMS

Juddson: In this episode of EMTalk, I had the pleasure of sitting down with my good friend and colleague, Adam Wester. Adam is a seasoned paramedic, paramedic instructor, and Assistant Program Director at Texas EMS School. Our friendship spans over a decade, beginning when I was training to become a paramedic. It’s been an incredible journey, from working together at Metro Care to collaborating in the educational field. 

As we catch up, one topic that’s been on everyone’s mind lately is the impact of Artificial Intelligence (AI) in EMS. This subject isn’t just a passing thought—it’s shaping the future of emergency medical services and will continue to do so in increasingly significant ways. 

Juddson: Well, here we are for another episode of EMTalk. And with me today is one of my good friends, a paramedic, paramedic instructor, and assistant program director for Texas EMS school. Adam Wester. Adam, how are you today? 

Adam: I’m good. How are you doing?  

Juddson: I’m doing fantastic. So, Adam and I have been friends or have known each other for probably 12 years.  I first met you while doing clinicals as I was becoming a paramedic, and then we worked together at Metro Care. 

At one point, we talked about how it would be cool if you came and worked at Texas EMS School, and here we are today.  

Adam: Yeah, it’s fantastic. We’ve both been on an exciting journey to get here. That’s for sure. Well, something that has been on my mind and everybody’s mind recently or lately is AI. 

Juddson: I want to give you a little time to talk about who you are and how you ended up in EMS. And then, I want us to roll into talking about AI and EMS because it will greatly impact how EMS works. I want to go into the trenches and figure out how weird it could get with AI, maybe how good it can get or how bad it could get. 

We could be talking about a Terminator situation. Who knows? We could also discuss how AI can solve your problem in a matter of seconds rather than minutes. So, Adam, how did you get into EMS?  

Adam: I got into EMS by accident. I was working as a corrections officer for the state of Texas and doing volunteer firefighting. 

After that, we had contract issues with the ambulance company in the area we covered. I was like, you know what? I’m just going to get my ACA and try to help. I went through that program and saw the potential that I had to be able to help people before the ambulance got there just by becoming an EMT. 

While in EMT school, it didn’t make sense to stop, so I went on to become a paramedic. A lot of it was because my dad was sick, so I wanted to be able to help him. So, that was a lot of it in the paramedic aspect. 

I’ve seen a lot of things; I’ve gotten to do a lot of cool things. I’ve seen a lot of technological changes throughout my career, just in the last 12 years. And it makes you think, where are we headed next? Oh, yes, I think about that a lot, especially since we’re on the education side now, because it makes a huge impact on what we teach these students. 

We must stay current because otherwise, we’re sending them out in the field with old information, which is how it feels in education. By the time the information changes, like a national standard level, it’s already been done for years. Typically, about 5 to 10 years.  

Juddson: By then, it’s changed again. And so, they must go back into the field and learn the next new thing already. Adam, you were in the Marine Corps?  

Adam: Yes, once a marine, always a marine.  

Juddson: Were you in law enforcement in the Marine Corps?  

Adam: No, I was infantry. My primary job was machine gunner. I got to play with all the big guns.  

Juddson: What is the biggest gun you ever shot?  

Adam: Mach 19 and 50 caliber.  Mach 19 is an automatic grenade launcher. 

Juddson: That’s so cool. So, I’ve got to shoot a 50-caliber sniper before, and it was, like, the most insane experience, and it put a massive hole through a refrigerator, like, 100 yards away. And I know, 100 yards is like chump change for shooting with that kind of gun. But it was fantastic. 

Adam: I like the Mach 19; you do not have to be accurate.  

Juddson: Adam, you have a wife and two kids.  

Adam: I have been married for 20 years now, and I’ve got an adult son in college and another son in high school this year. 

Juddson: So, you and I sat down to figure out a topic for the show. You and I talk all the time; we can talk about anything. This morning, I was using AI to help me think of things we could talk about a little more deeply. 

As I sat there, I said, that’s it. That’s the thing: AI.  We both work for Texas EMS School, which sponsors our podcast, and when we try to devise wording for policies and procedures, we often use AI. What’s the next step in the medical world? 

It’s already being used in some regards, but the way that we use AI now is more to help us with quick searches or things like that. But I want to figure out what our limits are. Do we think that it could get worse the way that we use it? 

So, first question: do you use AI for anything right now?  

Adam: Currently, the only thing I use AI for is to bounce ideas off if I don’t want to bother someone I like.  

Juddson: Why would we teach people the same way anymore when technology is right there at their fingertips? You know, we grew up thinking or being told you won’t always have a calculator. And then, five years later, we always had a calculator.  

The entire world is changing in terms of how we go about that. Using AI, we can simplify education and get people to start thinking critically instead of saying, here, memorize all this information, right? 

Adam: We can recall the information with AI and then do the critical thinking part.  

Juddson: When does the critical thinking come into play? Can we do that in the ambulance?  

Adam: I think that is a slippery slope, honestly. You still must have the base knowledge of how the body works, what signs and symptoms you are seeing, and a general idea of what you are looking at with your patient. 

Eventually, AI will get good enough to utilize a camera, and AI will diagnose.   

Juddson: So, you’re thinking more along the lines of we’re giving input and information, and then AI helps us come up with those differential diagnoses. 

Adam: Yes, we’re close to that. 

Judson: If you have an AI program already helping you write your narratives, then you’ll already have some training in the AI model. It will know what keywords you’re looking for, such as signs and symptoms. It will know what blood pressure, heart rate, respiratory rate, and other things. It will say, okay, humans have said that these signs and symptoms equate to this diagnosis. 

AI can predict quickly, reducing the time it takes to figure out a treatment plan. That’s how I see it working at a considerably basic level. I consider when AI lets me decide and whether it is a good idea. 

Adam: There would be a checks and balance system, and I can’t imagine a world where people say, I will take all the liability for that decision. I can’t imagine a world where any developer would say, “My AI model is so great that we’ll take the liability for any medical malpractice.” 

Juddson: For example, I can see the AI model as “You have uncontrolled bleeding; it’s above the neck. Go ahead and put a tourniquet on.” How simple would it be to write the algorithm in a way that recognizes that you can’t use a tourniquet if it’s above a specific body location? You can only use pressure. I think of things like that all the time because we create algorithms that tell us, you know, what we should do in those situations.  

So, if you build those algorithms, they become a tool to help you confirm what you think is the right thing to do. 

Adam: Maybe that’s where we determine what’s realistic. But if we took that further and are like, what’s unrealistic? And I wonder if these things are realistic because technology is advancing fast. I heard recently that they’re thinking of ambulances driven by AI systems. 

This concept has already been tested in the civilian market with different newer models of self-driving or self-piloting vehicles. I know there are taxi services that already have driverless vehicles. 

And that’s mostly over in California, where they’re testing this stuff out. It’s possible to take that same technology and put it into an ambulance.  I saw an article the other day where they put it into an 18-wheeler, and a carrier will start testing 18-wheelers with automated driving. 

Juddson: I mean, think about that. You have thousands of pounds just barreling down the road, and there is no human to intervene. But most accidents are caused by human error here. I get that, but it is still a scary thought. 

Adam: If they can get it to work with an 18-wheeler, they can get it to work with fire engines, and they can get it to work with an ambulance.  

Juddson: What about those situations where we drive with due regard but bend laws a little bit because of an emergency? How does the system know that it’s okay to do that? 

Adam: I assume the same way it knows everything else. An algorithm portion would say, “We’re in an emergency mode based on a call received to whatever address that’s geocoded.” Then, it figures out a path and takes that path. It would have to include some automation system as well. 

Judson: It would be a game changer because then you could always have two providers in the back.  

Adam: Exactly, and it makes it where you can continue transporting. 

We can get them to the hospital faster and provide a higher level of care. I do think that we’re headed that way.  

Juddson: It’s going to be interesting to see. The first time you and I watch an ambulance go by with nobody driving, it will be super weird. I really do, and I think we’ll see it in our lifetime.  

Adam: Oh, I’m sure, in the next 5 to 10 years, they’ll have an ambulance with the technology to drive itself to a call and then to the hospital.  

Juddson: What if we gave AI the ability to make physician-level decisions in an ambulance? What if it became our med control? That’s an interesting concept. I worked at a service that put in a camera system directly linked to a level-one trauma facility. 

If they needed to, they could call this level-one facility and say, “I need a neurologist, or I need a trauma specialist.” They could talk directly to that physician, who could see everything in the back of that ambulance. This gives the ability to assess the patient. 

Adam: Based on that system, it wouldn’t be too far-fetched to have an AI model do the same thing. Maybe have a doctor look at it on the other end and say, “Yes, this is appropriate.” A doctor would verify the AI’s decision, but the AI system would be built with physician-level algorithms instead of paramedic-level algorithms. 

Juddson: Really, the solution is to train paramedics better so that they can continue to increase the decisions they’re capable of making. But that’s a more extended conversation. Much of that is about legislation and trust. 

Adam: Yes, I think, unfortunately. And then there’s the medical community, you’re closer to trusting a machine to do what they do than they are trusting us to do what they do. 

Juddson: It depends on the physician. Doctor Philpott, our Medical Director for Texas EMS School, is one of those who tries to advance it and give more ability. 

Even starting the first EMS program was an uphill battle, and now we keep pushing it. However, AI may help us fight that battle because it may be considered a safeguard, such as double-checking ourselves with AI, or it could be on the opposing side. 

We were talking about education earlier in this podcast and the fact that we need to focus less on memorization of information and more on teaching people how to use a resource to find that information and then use critical thinking to utilize the information. 

Adam: That concept is the same within EMS. Take, for instance, medicines; there are a million different drugs out there, and it’s great if we can memorize them. However, if AI comes into play in EMS, it could assist in such a way as to say, “Grab a vial of this; you’re going to draw that up. Here’s what it’s going to do. Here are the side effects. Here’s what allergies you need to be worried about.” That gives us the information.  

Judson: I went on a call once, and this person had a cabinet full of meds that they take. Most of the time, if we don’t know a medication or what it does, we look it up and Google it, right? It would increase the speed at which we do that.  

Adam: Essentially, AI programs work like high-level search engines that can produce results and organize them rather than just producing the results. They even go as far as explaining the results of what they found, which could make a dramatic difference in our speed of being able to deliver the care we are giving.   

However, AI can look at the medications and tell if any of these will interact poorly.   AI then looks at the meds and looks them all up. Granted, who knows if it saw everything? That’s the thing we’re talking about here is the fear of what if AI screws up and it tells you, “There’s no negative interaction.” 

Juddson: That would be interesting. You mentioned hiccups or glitches.  

Adam: Yes, that might occur. It’s going to happen. I mean, there’s going to have to be a whole separate insurance for the use of AI technologies.  

Unknown 

Juddson: Some concepts like that could be interesting, and then we could take them to the extreme. This is where I want to start ending this conversation. What if we no longer need providers?  

Adam: You’ll always need providers of some kind. 

There’s physical lifting. The actual physical skills need to be applied to your patient. So, whether there is intelligence like what we currently have in our system would be debatable if they used a ton of AI.  

Juddson: They do the lifting, they start the IVs, that kind of thing, but otherwise they’re just technicians. What if it took us backward in our profession, and we just became technicians again, and that’s all we were? Is that a possibility?  

Adam: I mean, if you want to think about it negatively. Playing devil’s advocate here could be, one of those things is just, you know, currently, our job description, if you think about it, is an emergency medical technician. 

Juddson: That’s what we are. We’re technicians. Right. Just with a whole lot of critical thinking that goes into it.  

Adam: If I were to take the critical thinking out, yes, it would be just a technician job. Where you go in, you do what the AI model tells you to and take the patient to the hospital. And that’s the scariest thought: if we advanced to the point where we put ourselves out of a job, there’s tons empty job. 

Juddson: Part of health care is comforting people. I put a lot of effort into being calm and making them feel relaxed, even if I think they’re going to die when I run EMS calls. I’m honest with my patients when I feel that way, but I’m also very calm about it. That’s true of most EMS professionals. That’s what we’re truly good at, the empathy and sympathy aspects. You can’t give sympathy and empathy to an AI model. You can’t teach it genuine emotion. 

Have you ever watched all the Terminator movies? I mean, that was the whole concept. They could make it look like a human, but they could never make it act like a human. Right? Because you can’t, you can’t fake compassion. Well, I guess a bunch of sociopaths out there would say differently, but you can’t teach a machine how that works. 

Adam: In 5 to 10 years, a model might be out there that can synthesize human emotion.  

Juddson: We’re not producing the highest-quality AI content, but we certainly have the highest-quality instructors. You could get precisely what you’re looking for through us. So, thank you, Texas EMS school. Ensure you go to our website and check us out, texasemsschool.com. Give us a call anytime at (325) 218-4444. And we will see you all next time on EMTalk. 

EMS professionals must navigate this rapidly evolving landscape with caution and curiosity. Integrating AI into EMS could lead to unprecedented improvements in patient care, but it also presents challenges that we must address head-on. Whether AI becomes a formidable ally or something more unpredictable remains to be seen—but one thing is sure: the future of EMS is set to change dramatically.