Maggie McKay (Host): Welcome to Conversations Like No Other, presented by Valley Health System in Paramus, New Jersey. Our podcast goes beyond broad everyday health topics to discuss very real and very specific subjects impacting men, women, and children. We think you'll enjoy our fresh take. Thank you for listening. I'm your host, Maggie McKay.

Today, we have with us Dr. Matthew Sumicad, emergency medicine physician, to discuss AI in the ED. Welcome. Thank you for being here.

Matthew Sumicad, MD: Thanks so much for having me.

Host: Absolutely. So for someone who is not familiar with AI in healthcare, what does AI look like in an emergency department?

Matthew Sumicad, MD: AI takes several forms in the emergency department. I would say there's really three main ways that we use AI in the emergency department, especially here at Valley where I work. The first two ways are kind of tied together, and they both have to do with writing the medical record. We've got passive listening devices. So, these are apps you can put on your phone that will sit in the room and passively listen to the conversation that you have with the patient. And then, afterwards, you'll kind of add any additional information you want, and it'll synthesize that information into a medical record.

You know, when we have a conversation with the patient, it's usually a little bit more informal, and then, you know, when we write our medical chart, it says things like the patient is a 35-year-old male, you know, and he's coming in with chest pain. The way that it turns a conversation into an actual medical record is very impressive when it works and very useful and time-saving.

Another way that we use AI is the note-writing. If we're not using these passive listening devices, we can kind of put the information that we have as part of our medical record into the note or put our note into a AI scribe, and it'll synthesize the information into a more readable and more linear format. This can be really useful in terms of saving time and, you know, increasing the efficiency, of course, of the department.

A final way and one of the ways we use here at Valley is actually AI imaging interpretation. So, we use an app like Viz.ai, but there's several others out there which will look at the images for CT heads or CT scans of the body and kind of flag them as abnormal even before the radiologist sees them.

There's a lot more areas that people use AI in the emergency department. But, you know, those are some of the major ones. I am looking forward to seeing how more it can get implemented in the department that we work in.

Host: Where are we currently seeing AI make the biggest difference in emergency care?

Matthew Sumicad, MD: Definitely, it's going to be with the reducing administrative and documentation burden. We spend up to a third of our time writing the chart, this can really reduce that time that's kind of not spent directly interacting with the patients or, you know, managing them in productive ways. AI scribing and the passive listening devices that I mentioned really can help reduce those kind of unnecessary time and increase the efficiency of the department.

There's other ways out there. But again, I think this is probably the most impactful way that we have in terms of actually using AI in the emergency department.

Host: Can AI help make the ED more efficient and reduce wait times? Because at least in medical shows, you always see the wait times. It's like piling up people.

Matthew Sumicad, MD: This is everybody's favorite question about the ER. It doesn't matter what intervention it is. Everybody wants to know, can we reduce the wait times? And see, nobody can wait anymore, you know? And yes, of course, it can be implemented in some ways. You know, AI triage is a way. I had mentioned the note writing can reduce some of the time spent computing.

Although I'll, say that I've looked into some of the data here. And actually, when the amount of time saved used by AI is kind of tracked, it doesn't actually reflect a significant time duration in the wait times because doctors end up giving that time back to the patients to have a longer conversation about the results or the findings.

And so, the time is actually used more efficiently and used more productively in the patient encounter than just shorter visits, shorter times, shorter wait times.

Host: So, does using AI change the role of emergency department staff or does it simply just provide an additional tool?

Matthew Sumicad, MD: I know this is a huge thing. I myself have worried about just being replaced by a little laptop with a question and answer microphone. But the best tools are the ones that you actually don't realize you're using, right? The computer was once a tool that was brought into the healthcare, and now it's essential, right?

The microphone for dictation and ultrasound machine, these tools are now just integrated into the way that we practice medicine, and we almost couldn't envision doing medicine without these objects. You know, the roles of the staff in the emergency department from, you know, doctor to triage to nurse assistant, APP, I think these roles will still be essential and still pretty much be there.

Because while the AI can help in writing the note and understanding who's sick, it can't get you a bed. You know, it can't change you. It cannot push medications. It can't shock your heart. It certainly can't save your life. So, our jobs are easier from the tools that we've developed, but they won't really just be replaced by these new tools out there.

Host: Good point. Especially when you're not feeling well or you're in a trauma situation, you need a person for sure, and the help of AI. So, let's talk about the patients and AI. How can a patient use AI to make their ER experience easier and smoother and more efficient?

Matthew Sumicad, MD: So, I think actually the best way that patients can use AI is outside of the emergency department. Understanding their own medical history, understanding their symptoms and their context of, you know, what they're feeling, and being able to summarize their symptoms so that when they do end up going to the emergency room, they're a little bit more concise, and they can give us the information that we need.

You know, when we speak to patients, we have kind of a mental checklist sometimes in the back of our head of we have to answer certain questions and know whether there are certain diagnoses that we should be worried about. And if you're the patient, you don't know what these are, and AI can kind of help you understand as a patient what it is that we're looking for. And so, we can be on the same page.

I think AI is also helpful kind of towards the end of the encounter when you receive a diagnosis or you receive some findings, AI can also help you understand, you know, medical terminology, which is there is a lot of complicated medical terminology out there. So, AI can really help understand what we're talking about and really be a great bridge to explaining medical terms, to understanding exactly what you're experiencing and what this means for your health.

Host: Yeah, use it to your advantage, right? How can we make sure that as healthcare becomes smarter, it also stays human? As we were mentioning, you still need a human to do all those things you listed.

Matthew Sumicad, MD: Yeah. And I think that this is a paradox, of course, because we want to live in a more technologically advanced age. And so, how do we still maintain some of the human aspects that make medicine kind of fundamentally a human science? And, you know, one of the major ways is really spending more time with our patients.

As I mentioned before some of the data had shown that even when the documentation time is reduced, the visit time isn't that much reduced because doctors end up spending more time. They have more time to spend. I always try to educate my patients, you know, when it's time to send them home.

I try to spend a lot of time having a conversation, making sure they understand the results, making sure that they understand the uncertainty of medicine. That is sometimes the mix of art and science, you know, where you don't always have exact answers, but you have kind of an understanding of things overall.

It's a difficult thing to answer at the end, but it has to do with the person and just trying to be more human, you know? Making sure you're looking them in the eye, making sure you're not sitting there on your phone or on your computer while you're interviewing them. Just simple things I think will end up going a long way when we end up becoming more and more plugged in as time goes on.

Host: Isn't that the truth? I mean, it's such a simple concept, but a lot of doctors do that. They're looking down because they're concentrating. I get it. But it's like, "Hello, I'm here. Look up."

Matthew Sumicad, MD: Yep. I mean, many times I've been like interrupted, you know. You end up running out of the room to take care of something in the middle of the conversation, you know. They call you overhead. You get a phone call from some other, you know, radiologist or some sort of finding. So, it's hard to focus in on the moment sometimes.

Host: is there anything else in closing that you'd like to add that you'd like people to know?

Matthew Sumicad, MD: People should not be scared of AI in healthcare. I think people think it might take away some of the human elements where, you know, you're just plugging your symptoms into a chatbot. And then, I'm talking to a chatbot. And then, eventually, we'll just have two chatbots talking to each other and no people involved at all, right?

But I think, you know, that's kind of a dark future that I don't think is really the direction we're headed. I think really we're headed to more of a place where we are using AI just like we would use, again, other tools in healthcare and in medicine to help make our experience smoother, easier, and more understanding.

Because we have things that we're looking for in healthcare, and we have kind of our own language in healthcare, the medical terminology and the medical understanding, and being able to communicate that to our patients, I think AI is essential moving forward in terms of making that an easy transition and making that understanding from the person who doesn't know medicine to understanding our perspective as a medical professional. So, I think it's going to be a good bridge as opposed to just kind of a barrier between humans.

Host: That's very encouraging. Thank you so much for sharing your expertise. This has been such useful information and, again, very encouraging for people who are worried about AI in healthcare.

Matthew Sumicad, MD: I hope so.

Host: Again, that's Dr. Matthew Sumicad. To find out more, please visit valleyhealth.com/emergency. I'm Maggie McKay. Thanks for listening to Conversations Like No Other, presented by Valley Health System in Paramus, New Jersey.

For more information on today's topic or to be connected with today's guest, please email valleypodcast@valleyhealth.com