Physician and nursing shortages: How AI can help [PODCAST]

Listen and followApple PodcastsSpotifyYouTubeMore apps · Browse all episodes

In this episode, we talk with Harvey Castro, a physician and health care consultant, about how artificial intelligence (AI) can help tackle the physician and nursing shortages in the United States. We discuss the contributing factors behind the shortages, the ongoing efforts to address them, and the potential role of AI in enhancing diagnostic and treatment capabilities, streamlining administrative tasks, and improving medical education and training. Join us for an insightful conversation on the future of health care and how AI can play a pivotal role in improving the overall efficiency of the health care system.

Harvey Castro is a physician, health care consultant, and serial entrepreneur with extensive experience in the health care industry. He can be reached on his website, harveycastromd.info, Twitter @HarveycastroMD, Facebook, Instagram, and YouTube. He is the author of ChatGPT and Healthcare: The Key To The New Future of Medicine, ChatGPT and Healthcare: Unlocking The Potential Of Patient Empowerment, Revolutionize Your Health and Fitness with ChatGPT’s Modern Weight Loss Hacks, and Success Reinvention.

He shares his story and discusses his KevinMD article, “Tackling the health care crisis with artificial intelligence: Combating physician and nursing shortages in the United States.”

The Podcast by KevinMD is brought to you by the Nuance Dragon Ambient eXperience.

With a growing physician shortage, increasing burnout, and declining patient satisfaction, a dramatic change is needed to make health care more efficient and effective and bring back the joy of practicing medicine. AI-driven ambient clinical intelligence promises to help by revolutionizing patient and provider experiences with clinical documentation that writes itself.

The Nuance Dragon Ambient eXperience, or DAX for short, is a voice-enabled, ambient clinical intelligence solution that automatically captures patient encounters securely and accurately at the point of care. Physicians who use DAX have reported a 50 percent decrease in documentation time and a 70 percent reduction in feelings of burnout, and 83 percent of patients say their physician is more personable and conversational.

Rediscover the joy of medicine with clinical documentation that writes itself, all within the EHR.

VISIT SPONSOR → https://nuance.com/daxinaction

SUBSCRIBE TO THE PODCAST → https://kevinmd.com/podcast

RECOMMENDED BY KEVINMD → https://kevinmd.com/recommended

GET CME FOR THIS EPISODE → https://earnc.me/wDBEjW

Powered by CMEfy.

Transcript

Kevin Pho: Hi, and welcome to the show. Subscribe at KevinMD.com/podcast. Get CME for this episode by clicking on the CME link in the show notes. Today we welcome back Harvey Castro. He’s a physician and health care consultant. He’s the author of the book Revolutionize Your Health and Fitness with ChatGPT’s Modern Weight Loss Hacks. Today we’re going to talk about his most recent KevinMD article, “Tackling the health care crisis with artificial intelligence: Combating physician and nursing shortages in the United States.” Harvey, welcome back to the show.

Harvey Castro: Hey, thanks for having me. I appreciate it.

Kevin Pho: So we had you on recently talking about ChatGPT and AI, and you’re all over LinkedIn as the physician AI expert. Tell me, how did you get so interested in this facet of health care tech?

Harvey Castro: You know, honestly, it really boils down to the iPhone moment. I was working in the emergency room, and I saw how useful the iPhone was. This was when the iPhone 1 came out, and I came up with my first app, called IV Meds, and I was just fascinated. So fast-forward: ChatGPT came out in December, and I was playing with it, and I had that aha moment where I was like, “Wow, I really need to do something in health care.” And honestly, I looked at it as a public service announcement: Let me help doctors learn how to use this technology, the good, the bad, and the ugly. But obviously the patients come first, and I wanted to do the same for patients. What started as a PSA announcement became more, and so now I’m really into this.

Kevin Pho: Yeah, we’ve only been, what, four or five months into the ChatGPT revolution, and now we have Google Bard, we have GPT-4. It’s moving so fast. Has it been what you expected?

Harvey Castro: Honestly, no. Like I said, it was funny, because I really thought, “Oh, I’ll just do this and kind of put a book out and help.” But honestly, I mean, I knew it was kind of like the iPhone, but I didn’t think it was going to explode as fast as it did.

Kevin Pho: Perfect. And there are so many tangents we could talk about at that intersection between AI and health, but we’re going to focus on one particular tangent from your KevinMD article, about physician and nursing shortages. Now, before we get into how AI can help, give us some context. Tell us, why is this a crisis?

Harvey Castro: Yeah. Unfortunately, as you and I know, a lot of great colleagues are retiring early. A lot of doctors are burned out, and it’s a combination of things. As our population is aging, there are more patients to be seen. Obviously, post-COVID, there’s a lot of strain on our health care system. And ironically, even though we’re not having the COVID breakout right now, a lot of patients are asking for a different type of health care. And as a result, I know personally that a lot of nurses are also quitting.

So just to give you a quick stat: According to the Association of American Medical Colleges, by the year 2033, the shortage will be roughly between 54,000 and 140,000 physicians. And then as far as nursing, according to the American Nurses Association, they project over a million nurses that we still need as of the year 2022. So we’re really behind the eight ball here.

And so I started looking at this as a problem, and then I started freaking out for my personal health. I thought, “Huh, if we’re at a shortage today, and we’re projected to get worse, and we weren’t really accounting for COVID and all these other things that are going on, what will our health care be in 10 years, 15 years? Is it going to be worse than it is today?” And so I thought, “Well, we need to come up with some solutions.” I know our government is adding more doctors. In the year 2020, they’re allowing more rural physicians, and there are more grants to pay for residencies, and they’re trying to alleviate that as an issue. But I still worry that this is not going to be enough for what’s going on.

Kevin Pho: Yeah. So before getting to solutions, just give us some stories. You’re an emergency physician.

Harvey Castro: Yes, sir.

Kevin Pho: So what are some stories that you’re hearing from your colleagues, and from the other doctors and nurses that you talk to, about what the situation is like, specifically in emergency medicine?

Harvey Castro: Yeah, it’s sad. I’m going to go back to the peak of the pandemic. My colleagues were saying they couldn’t keep up. Unfortunately, they were having the nurses do as much as they could in order to get as much done, making clinical decisions that technically the nurse shouldn’t be doing. But there was just such a mass of patients coming in that they literally had people dying in the waiting room. And so it was just an overabundance.

Today, I’m hearing similar things, in the sense that they’re just seeing so many patients and there are not enough doctors, and they’re having to rely more on PAs. And from a lot of doctors, especially in the emergency room, I’m hearing a lot of burnout, to the point where, about two years ago, I got my MBA, and I want to say 70 percent of my class was emergency room physicians.

Kevin Pho: Wow. And I’m sure you heard the story of the recent Match. There are, like, 500-plus unfilled spots in emergency medicine.

Harvey Castro: Yeah, which is a shocker, because when I went through, I felt so lucky to be able to get an ER spot, and I know it was really tough to get in. And when I saw that, it made me sad for our profession, because I thought, “If the future doctors of America do not want to be in the emergency room, what’s going to happen in the ER?” You know, it’s scary.

Kevin Pho: All right. So you’re an expert in that intersection between artificial intelligence and health care. What are some ways that AI can help alleviate these shortages?

Harvey Castro: Well, for one, it’s been accepted in radiology, obviously. The latest technology that I heard about that I thought was exciting is this: You see a patient in the ER, and unfortunately they’re having stroke-like symptoms. You get a CT angio, and an AI is able to look at the CT scan, analyze it, and then, before the radiologist even sees it, send you a text to say, “Hey, this person may have a stroke,” or “has a blockage,” or whatnot. And I thought, “Man, this is amazing.” Instead of waiting on the radiologist, who has this big stack of CTs, now you’re getting a quick text saying, “Hey, this was read by the AI.” You still need a human to confirm it, but that’s amazing.

Unfortunately, our culture obviously doesn’t accept change. Change has been gradual. In the fields of radiology and pathology, in these fields, we accept AI pretty readily; we think it’s OK. Whereas right now, ChatGPT came on so strong, and so many changes have been made, that there’s a lot of pause. And so I totally understand. My understanding and goal with ChatGPT is to honestly recognize that there are hallucinations, and recognize that there are issues with ChatGPT, things it cannot be used for today.

Kevin Pho: So when you say hallucinations, that’s a specific AI-related term. For those who aren’t familiar with that, what does that mean?

Harvey Castro: Yes. So to explain hallucination: Think of it as a statement that the computer makes, and that statement basically could be half true and half false. And as the unknowing reader, you may look at it and say, “Oh, that’s the truth, and I’m going to use this technology,” when in reality you shouldn’t.

And so my analogy is: If you’re an ER doctor, you can see other ER literature and say, “Yep, this is correct.” But if I look at literature from another specialty, I’m just going to say pediatric cardiology, and I’m going through it, and I’m like, “Yeah, this looks good,” but then I show it to a pediatric cardiologist, they’ll say, “Oh, Harvey, no, no. This is so wrong. This is not even close to correct.” As a doctor, I can bless only things that are in my field, which is ER.

Kevin Pho: Excellent. So even though the AI technology is not perfect, you still feel it can help?

Harvey Castro: Yeah. So some of the things I think it’ll help with today are in the medical school arena. Obviously, I feel like doctors in medical training are good enough to understand what’s right and what’s wrong, and they can use it for clinical vignettes. They can use ChatGPT to create flash cards, to quiz them, to test them.

As that technology gets better, and I’m not advocating straight ChatGPT, I believe in the future we’ll have a health care-related, let’s call it BioGPT, or some health care database that’ll be very specific to health care. And when we have that tool, then I think we’ll be able to do more advanced applications. But as of today, I think it can help with simple things. Before, we had to schedule our patients and all that; all that’s now being automated with chatbots, and now with ChatGPT it’s being taken to another level. So I think that’s going to help. Ultimately, my goal is, if we can become more efficient doctors, then we have more time with our patients.

Kevin Pho: So take us through a walk in the emergency department. If you were in charge, specifically, how would AI and chatbots help the daily workflow in the emergency department?

Harvey Castro: Yeah, I’m going to blow you away with this one. I would have a camera that takes your blood pressure as soon as you walk in. It takes your blood pressure, your temperature; you would have your vitals. It would go and be fed into an algorithm. When you walk in, your voice would also be recognized, so that it knows if you’re depressed, if you’re having early Alzheimer’s; they would already know. All of that would go into a system, which would go into a ChatGPT equivalent and would triage the patient. That patient would immediately go into the room and be categorized, and then immediately we would know what labs need to be done.

There would be no charting. You would go in and have a one-on-one conversation, and there’d be two AIs going: one capturing our conversation and helping me document and helping me bill it, the other making sure that everything’s accurate, but also putting it in categories, so that when we finish our conversation, I can get a risk stratification: “I may have not asked X and Y, and these questions are important.”

Another thing in medicine that I think is important, especially in the ER, is a differential diagnosis. If I don’t consider certain things, it’d be nice if ChatGPT or an equivalent would say, “Hey, there’s a 1 percent chance of this disease. You haven’t asked these questions. Make sure you ask.” And then I would ask. These are all things that I think will be possible. That’s how I see AI. I know that’s an extreme example, but I do see that happening.

One other thing I usually like to throw in is that in the future, we’ll be able to print our own medications. So I do see that as another platform, where I write a prescription, and the patient, and this is actually already out there, can print their own medications. But what I’m getting at is that there’s a new technology where, whenever you use the restroom, there’s an application that can look and see what the metabolites are, and therefore the doctor can readjust that medicine in real time. So instead of going back to the pharmacy, they can do that for their patient.

Kevin Pho: Now, you’re framing this technology as a way to help with clinician and nursing shortages. Is there a possibility of this technology working too well and eventually taking the jobs of some clinicians and nurses?

Harvey Castro: That’s an excellent point, and I know that’s a big fear. I personally don’t see that. I just feel like there’s so much hands-on, and there’s that physician gestalt and that physician touch on patients, saying, “Hey, everything’s going to be OK.” That will never be replicated by a computer or a robot. I think it’s just an augmentation. It’ll be an assistant. Just like I have my smartphone, I think it’ll just be another tool that I can have, other than my stethoscope. I’ll have a GPT equivalent to just help me with my patient care.

Kevin Pho: So what about trust in AI? Because we always hear about these edge cases, about how AI sometimes gets it completely wrong, and some outlier cases, and inevitably these types of mistakes get a lot of publicity and decrease the trust in AI. What will it take for the medical establishment to trust these AI tools?

Harvey Castro: I think ultimately it’s going to have to be a database that is used the way everybody uses UpToDate. If there was an UpToDate trusted source that became a GPT equivalent, if you used PubMed and all the resources, and you just knew that this was validated by all these sources, I think at that point it would give us that trust.

The opposite effect that will happen is, I’m worried that some physicians may become too reliant on it, and they’re going to forget their tools: the simple physical exam, touching the patient, some of the basics. They’re going to let it go and start relying more on AI.

Kevin Pho: We kind of see that already with the advent of technology and scans, and the amount of technology we already have at our disposal. Some are saying that the physical exam is almost becoming antiquated.

Harvey Castro: Yes, yes, I’ve heard that many, many times. But I’m able to pick up things pretty quickly with a good physical exam. I think it’s becoming a lost art. But I love talking to doctors in third-world countries that don’t have access to this technology, and what an amazing exam they’re able to do.

Kevin Pho: So like we talked about earlier, the trajectory of AI is going so fast. ChatGPT was literally only announced at, what, the end of November or something like that?

Harvey Castro: Yes, sir.

Kevin Pho: What needs to be done to keep it on track, to prevent it from going off the rails? What kind of red flags do you anticipate going forward that we need to avoid in order for AI to achieve its true potential in health care?

Harvey Castro: That’s a tough one. I think the main thing is always verifying, not taking everything that it says as valid. Always verify everything, even if ChatGPT gives you references, because you can ask it for references now. I would verify those. My biggest thing that I’m trying to do with patients and health care is teach them how to use this tool correctly. Ironically, some doctors are using it incorrectly, and the way they should be using it, they’re not using it.

Kevin Pho: Give us some examples of that.

Harvey Castro: Yeah, so for example, they’re not sure about the hallucination effect, and so they’re using it. For example, I’m an ER doctor, but I may be using it in a field, let’s just say pediatric cardiology, and asking it specific questions and saying, “OK, this must be true. I don’t know it, but it sounds right.” And that’s not the way to use it. It’s to use it to reinforce maybe some of the areas that you’re already good at. So as an ER doctor, I’m asking it for a differential diagnosis, but then maybe asking it about some of the rare diseases, where I’m like, “OK, I did think of X, Y, and Z.” I’m augmenting, instead of going into something totally blind. I’m using it in that way. So that would be an example.

Kevin Pho: We’re talking to Harvey Castro. He’s a physician and health care consultant. He’s the author of the book Revolutionize Your Health and Fitness with ChatGPT’s Modern Weight Loss Hacks. Harvey, tell us some of your take-home messages that you want to leave with the KevinMD audience.

Harvey Castro: Honestly, I want to thank everyone for your time. And really, just educate yourself on AI. It’s a new tool. Medical schools are not getting educated on AI that much, and I think as physicians, we’re lifelong learners. So make sure you’re making time to learn about AI.

Kevin Pho: Harvey, thanks again for coming back on the show and sharing your time and insight.

Harvey Castro: Thank you.

Listen and followApple PodcastsSpotifyYouTubeMore apps · Browse all episodes