Using AI to Become a Better Advocate for Your Health

share

Want information and notifications about upcoming podcast episodes?
Subscribe to IG Living Magazine.

If you have any concerns, comments or questions,
please email editor@igliving.com or call (800) 843-7477 ext. 1366.

Episode 22 Transcript

Using AI to Become a Better Advocate for Your Health

---

Welcome to today’s podcast. My name is Abbie Cornett, and I am the patient advocate for IG Living magazine. This podcast is brought to you by IG Living to give readers the chance to hear directly from healthcare experts on topics that matter to them most.

Abbie: 0:05
Welcome back to the IG Living Advocate Podcast. I'm your host, Abbie Cornett. One of the things I hear from patients all of the time is that managing chronic illness can feel like a full-time job. Between doctor's appointments, medication, lab work, insurance paperwork and trying to keep up with the latest information on their condition, it can feel overwhelming. On top of that, most medical appointments are very short. You may have to wait weeks or even months to see your healthcare provider only to realize afterward that you forgot to ask an important question or you didn't fully understand everything that was discussed. That's why today I'm excited for our episode. Today's episode is on artificial intelligence, or AI, and how it has become part of our everyday lives. While many people use it to write emails, plan vacations or answer everyday questions, it can also be an incredibly helpful tool for our patients.

When used responsibly, AI can help you prepare for doctor's appointments, organize your medical information, better understand complex health topics and even help you think through questions to ask your healthcare team. Now, to be very clear from the start, AI is not a replacement for your healthcare provider, and it should never be used to diagnose or treat a medical condition. But it can be another tool to help you become more informed, organized and confident and to advocate for your own healthcare. Today, we're talking about practical ways patients can use AI to do just that. I'm excited today for our guest, Stephen Walters. Stephen is one of our contributors to IG Living magazine, where he writes a column titled The Digital Patient, where he discusses practical ways patients can use AI to better manage their healthcare and advocate for themselves.

If you've read his articles in IG Living, you know he has a talent for making a complex topic understandable and may give you practical tips to understand things. Stephen, thank you so much for joining me today. Before we get started on today's topic, I'd love for our listeners to get to know you better and understand how you got interested in this topic. Can you tell us a little bit about your background and how you became interested in AI? And what inspired you to combine AI with patient advocacy?

Stephen: 2:54
Yeah, of course, and thanks for having me, Abbie. It's a pleasure. I've been looking forward to doing this since you mentioned it. And just to give you a little bit of context, so I do have a rare disease, and I'm also an IG Living contributor by writing The Digital Patient. And I do live in Canada, so I live just outside Toronto, Ontario.

And so to give you a little bit of context on my career and where I came into the AI space was, the past 12 years I was in pharma. So I was working sales, the majority of that was at Pfizer and commercial-facing. So, I was always in front of customers. I was on the road, having sales dialogue and conversations, but I've always had an interest in data and analytics. And what that meant was, working for a large pharmaceutical company and industry, we have a ton of information, a ton of data. And so I always found it interesting being able to get insights from that data, trying to uncover new opportunities or just gaps in my territory, because you're really in charge of running your own plan. And then, I feel like I've been an early adopter to AI just based off how early I was when some of the beta users were being released for Chat GPT and open AI. And I was just really mind-blown by how much these tools could remove a lot of the manual friction that I was doing on a daily basis.

And an example would be the spreadsheet work. I'm not technical by any means. I never have been. I've always been more of a salesperson. But being able to synthesize large amounts of information was something that AI was really good at. And that took me hours by making pivot tables or using Excel and kind of crunching the numbers within that. So, I think when I left Pharma back in March of 2025 to go into business for myself, at that point we were already a couple of years deep with a lot of these tools. And I just saw from the get go how much advancement there was in the AI space and how powerful these models were.

And I saw it from both a commercial lens and a lived experience lens as a patient. And it's helped me tremendously in both aspects of my life for just navigating my disease and then also just in my day-to-day work. So that's really a little bit about me — an industry pharma background that turned into entrepreneur and rare disease patient.

Abbie: 5:42
You know, I think your background is really unique, particularly having come from pharma and having a rare disease, that you understand how complex some of these issues can be. This is a really basic question because a lot of our patients are not familiar with it. What is AI? It can sound intimidating if you've never used it before. How would you explain AI to someone who has little or no experience with it? And why do you think it's becoming such a useful tool for patients, especially those living with chronic or rare diseases?

Stephen: 6:19
Yeah. And it's a really good question just because I think what I've learned is being in the trenches on a day-to-day basis, understanding all these new tools, these new models, it can get overwhelming. And there needs to be like almost an AI literacy where it's just helping people that have never really used anything in the AI space. And when you think AI, most people just think it's a chatbot with GPT, or the LLMs, as we phrase it.

But it's so much more than that. And not to sound too technical or get too technical, but I think, where patients could really leverage this now is large amounts of information that you used to use Google for. And I know the term was called Dr. Google, because you're oftentimes looking at potential therapeutic areas, disease states, medications that are available. You're searching online where these tools, these AI chatbots, these AI platforms can really do a great job doing research or due diligence on a specific niche area. So I think if, you know, I'm going to summarize this quite broadly, but it's a lot more than just going into GPT and typing a sentence and trying to get a response. A lot of it is around basic AI literacy, understanding what to ask, how to phrase that ask, how to build something called a prompt.

And a prompt is what you're actually putting into that search box because that dramatically alters what's going to come back when you're able to feed it good information. And I spoke to that in one of my articles on just what a good prompt would look like compared to something that was very basic and the difference of the output. So I think there's a number of different tools, and we can definitely speak to that further on, how to get started or where to start. But I think to just kind of summarize it very high level is that it's so much more than just thinking ChatGPT or an LLM.

Abbie: 8:23
Absolutely. and I know a lot of people have real fears about AI. And like anything else, it depends on how you use it. Like some of the ways that you mentioned, I use it all the time. I'll go to the AI, the different sites, and I'll ask it to pull articles that I can go to to actually read the source material. And it saves me a lot of time trying to find the source material.

One of the things I really appreciate about AI is it can help patients feel more prepared before they ever step into the exam room, like asking it to find articles on their diseases. One of the biggest frustrations that I do hear from patients is that appointments never seem long enough. And it's easy to forget symptoms that they've been having or questions that they wanted to ask. How would you recommend a patient can use AI to prepare for those doctor's appointments, to prioritize their questions and get the most out of that doctor's appointment in the limited time they have available with their healthcare provider? And I know this is a big question.

Stephen: 9:40
Yes, big question for sure, but happy to kind of chip at it and answer it because there are like use cases and examples that I have that I think would be valuable. So for myself, like I said, my I've always been obsessed with data and tracking, and I know that's not everybody, that's not every patient. But for myself, that only gets to see my immunologist once a year, I want to make sure that I'm using that 15, 20 minutes as best as possible.

And in my opinion, I think when you come prepared to a healthcare provider, you're taken a little bit more seriously. And if you can make the best use of that time, it can give you peace of mind. And I think that's the biggest thing that I've always said is you want to advocate for your health. And, so, I think to go back to that question you just asked me on AI and what does it mean.

You'd think again, just researching articles like you suggested, or using the chatbots to try to find medications or information. I used AI to actually build a tool that can house a lot of this information so I'm not scattered. An example of this would be, you're logging notes maybe in your phone on your app, or you're writing things down on a piece of paper. Having a centralized intelligence system where you're just feeding all this information, all your thoughts, all your research, all the comments from the doctor, even if it's something that you can give to a caregiver so that they're aware of your dosing, your medications, what you're feeling in terms of, you know, adverse events or anything like that, that can all be documented and stored. And I think that's a really good start for people just to have a centralized spot that they can go to and put their research findings, put their notes, put their questions.

And so in my example, I mean, obviously it's a little bit more technical, but I'm able to look at, on a day-to-day basis, how I'm feeling by a simple gauge of one to 10. I can pull in data from my sleep. I can pull in data from my heart rate and different apps that can all feed into it. But I think as a very basic start, having the right questions to go into a doctor's office could mean a lot because that could save time.

Usually the first questions that the doctor asks is: How have you been feeling? Well, great, I guess, or maybe not so great. And then they're gonna probably ask you some of the questions, like in my case, when was the last time you had an infection? When was the last time, you know you were hospitalized or something? But when I have this tool, all of that's already documented. So I can go in there saying, well, I feel OK today, but you know, two months ago on this exact date, I had a sinus infection. Here's the medication that was taken, and here's the duration that it was taken for, and here's how long it took me to feel better. So giving that context can then get fed into the doctor's end of notes, and it helps them make better informed decisions. Maybe that type of medication wasn't effective and they need to try something different next time. But having those types of insights when you go in and having them documented is how I formulate my questions when I go into these appointments.

Abbie: 12:59
You know, preparing ahead of time is really important. And, I've always told patients one of the first things that I always say is educate yourself about your disease. You're your best advocate. But I think we've all had that experience, particularly the newly diagnosed, where you leave the doctor's office and you're like, what did they just say? You know, what am I supposed to do? How can AI help make sense of some of that medical information and the terms that we hear? Patients often leave their appointments with pages of notes, medication changes, test results and like the terminology I mentioned. How can AI help the patient understand what the doctor said, the terminology? And how do they transmit that to their everyday lives and their caregivers?

Stephen: 13:54
Yeah, and I think that comes down to, and I think this is really applicable to caregivers too, especially with the elderly population. Oftentimes, they leave that appointment and it's all just being documented in their head, and it's really hard to remember what was said or what the next steps were or some of the medical jargon. Thankfully, based off my background, being in pharma and speaking to healthcare providers for over a decade, I can understand a lot of it, but I can definitely reason and respect. But many people can't. And so you want to be able to take as many notes as possible when you're in that appointment. Whether that's on your phone or if you're writing things down, but try to capture it in some sense because that's where AI could come in. And you can go and do the diligence afterward on looking at terminology that was said or any questions that you weren't able to answer in that appointment to go and do the research through these tools.

And there's a wealth of information. The LLMs, like GPT and Claude, will often pull up accurate information based off the sources and the citations that it's giving you that answer from. So it's kind of a tip for the listener to always look at the output and where it's coming from, because the different tools will pull from authoritative sources, mostly institutions that you want to look for. But I think the biggest thing is you need a way to capture that interaction. And, that's something that I'm still trying to explore and find a solution for. We, in Canada, have a lot of the offices now dictating the notes for the physician.

Abbie: 15:42
I was just saying you could probably take the dictated notes and use those. Because a lot of places are doing that here too.

Stephen: 15:45
Exactly. So that's where I'm thinking there should be something more on the patient side so that they can just focus on the interaction that they're having with the doctor and not have to worry and then get a transcript that's raw. And if you were to ever get something like that, of course, assuming that there's no compliance burdens or anything like that and the privacy and all that is respected.

But having a transcript would be so valuable because then you can just focus on the interaction. You can take anything that was said that was maybe just not understood at the time, or you're maybe overwhelmed and didn't think of to ask for clarification on the spot. But that would be really gold for an AI to take that transcript, dissect it and break it down. And that's where the context of a prompt would come in because you could say, listen, I am not a medical professional. I'm not asking you to give me any medical advice. I simply want to understand this as you know it. Explain it to me like I'm completely new or like I'm 12 years old type thing. And that's where it would it would shine because it could break down that language into something that's understandable.

Abbie: 16:57
Absolutely. And that goes back to the quality of the prompt. For many of our listeners, it's not just the one appointment. They've had their illness for a number of years, sometimes decades, and they have multiple providers. You know, I see patients with binders. And I mean, I used to have a binder years ago that had all the doctor's appointments, test results, yada, yada, yada, yada.

They're trying to manage all of this paperwork and test results and reports and medication and insurance paperwork. How can AI help patients organize all that information so it's easier to find and use when they need it?

Stephen: 17:45
Yeah, and that goes back to having that centralized system. And I call it like almost a brain where I'm housing all that locally on my computer. And the parameters of privacy and the safeguards could all be adjusted based on that patient's preference. But for my use case, I usually have everything stored in the tool that I built, my app.

And I can feed it lab results. I have my care team, which consists of immunology, internal medicine, rheumatologist, dermatologist. It can house all that information on when the last time I had them, what the notes were on that, what the next actions were, if anything, any imaging that was done, any blood work, so on and so forth. So that could be as simple as just creating something called a project, which a lot of these tools have the ability for so that you can just kind of house it right in there. Like I said, if you're nervous on the privacy side, then you can start running something called local models where only you're having access to it.

Abbie: 18:47
I was just going to ask about the privacy because a lot of the AI models are obviously not HIPAA-compliant.

Stephen: 18:55
Mm-hmm. Yeah, exactly. So, you know, anytime I'm feeding, and that's kind of why I built my own tool is to try to do something that's bespoke and custom to me. But some people are just kind of screenshotting some of their lab reports without putting any identifiable information in there. They can also turn off the settings for the models to train on their data so that's not getting reported back. There's a number of different ways that you can do it. But you know, for a very basic use case, you could still take a lab report or something and just try to house it and just see if there's differences throughout. You know, how often you're getting tested or if there's any trends or spikes or even just looking at a certain blood test and what they're testing for. Just searching for that and understanding what is the meaning of this number. Sure they give you a range, and if you're in that range then sure, you could be fine. But still, to try to understand it, you might want more information on what it means.

So, I would say again, trying to keep things organized in one centralized spot that you can always turn to and go back to so it's not scattered across different computers or different notepads or binders in your example. And then you can you can really leverage the technology that way.

Abbie: 20:19
Managing a chronic illness besides the doctor's appointments and insurance means keeping track of day-to-day things that can slip through the crack. When do I take my medication, reminders, refills, symptom tracking, which we covered a little bit, organizing a treatment schedule, and even keeping that running list of questions, which you briefly touched on when you talked about, running list of when you were on antibiotics and treatment. But, you know, symptoms, questions. What are some of your practical ways patients can use AI in their daily lives to improve their outcomes and stay more organized?

Stephen: 21:11
Yeah, and I think that's really good for looking at what's coming in the pipeline. For me, the medications in the Phase II and III studies are really interesting because, you know, I'm on medication for life, right? And just because it's dosed in a certain way or it might produce certain adverse events doesn't mean that there's nothing coming that might be better suited for me and my patient profile. So, I find that the best use case using AI is looking at what's currently being studied out there that's really hard to dig into, like PubMed and clinicaltrials.gov. Those are huge resources that you could turn to to just educate yourself a little bit more on what's available or what's been done.

Same with understanding recruitment sites, maybe you might be eligible to be in one of those clinical studies. So I find that really interesting, and that's something that would have taken me hours and hours before to manually look up and research, whereas now AI is really good at pulling that information in a short duration of time.

Abbie: 22:27
You know, that transitions really well to our next question. Because you mentioned some of the more scholarly sites, PubMed, and those are the sites that I use on a regular basis. But how can patients use AI to better understand the articles that are there and the new research? Or, how can they use it to stay up to date with their therapies and developments?

You know, because let's face it, you and I do this for a living, and we're used to the medical jargon and the terminology. But a lot of times the articles are, maybe not as easy to understand for someone that doesn't have a medical background.

Stephen: 23:15
Yeah, and I think without getting too technical or too complex, what I would suggest to anybody that's kind of new or entering the space is ask AI: How can I get the most out of learning more on my disease? Or, how can I leverage and tap into PubMed?

Abbie: 23:36
Would it help, would you agree, with using AI to summarize the articles when they do?

Stephen: 23:40
Yeah. Because, you can look at the abstracts and it can get quite confusing real fast. And so if you really just want to understand the studied population, where it was done, what the outcomes were, what the dosing was, what the medication was, that can all be broken down in summarizing that article. And Claude, for example, has made it really easy because they've had something called an MCP. Think of it as a plug-in or something that you can tap into so you can use the PubMed MCP in Claude that would give you access to that universe. And then you're just sparring and talking with Claude to say, OK, please turn to PubMed for this disease or therapeutic area; pull any recent articles in let's say the last two years on this disease state.

And once it does, you can then just go deeper and ask: OK, please summarize this, tell me how this would be relevant, what are the outcomes, what were the endpoints? And if you're not sure what an endpoint is, it'll explain it to you. So just using the chatbot to educate yourself, and if you're unsure, asking it how to explain it to you in a way that you can would be a huge win.

Abbie: 24:57
I was gonna say that's a really good idea. I haven't used it for that specifically. I was working on my degree in epidemiology and even some of the data in the articles in PubMed is just overwhelming. So I've used it before to consolidate, to say, hey, you know, what does this mean?

Of course, AI is incredibly helpful, and it's becoming more and more helpful for research every day, but we also know it's not perfect. And this is the one thing that you know I always do. Even if I have it summarize something, I still read the article. AI doesn't always get it right. You have what they call hallucinations, and it can provide incomplete or inaccurate information. How can patients recognize some of those red flags and verify what AI tells them to make sure they're using AI as a tool to support them and not replace them? And I 100 percent reiterate the beginning of the conversation: Do not use AI for health advice. Use it as a tool to learn. But when they receive advice, how do they how do they do that? Or how's the best way to protect themselves?

Stephen: 26:24
Yeah, and that's a good question, because hallucinations are a real thing, and you want to make sure. It's common. There's misinformation or it's not grabbing accurate data, and it could misinform you. And it's a space that, commercially, I'm looking at working with industry on: How are you ensuring that these drugs, these medications, these disease states are showing up relevant. And I think what you could do as a patient is really understand the sources and the citations that it's referencing. And, if I use the example of Claude, right, looking at after it does kind of an output to see where did it grab that from. It should be sourcing that information, and it should be linking it. And if you know if it's coming from something like the Mayo Clinic or an institution or PubMed, those are reputable and that's usually what the LLM's gonna try to pull from.

Abbie: 27:26
And I also actually check those articles to make sure that the links are valid.

Stephen: 27:35
Yes, you're gonna want to click on it. And they usually they have a direct link that you can then go in further and go right to the referenced URL. So yeah, looking at, OK great, it referenced this. I'm gonna go vet it and check it and do my diligence on it. But like you said, it's there to support and educate. It's never there to diagnose. And I know the models are cracking down on that. One thing I will say is, you know, throughout the last six years, and it's incredible how fast and how good these models are getting. So they are getting better and better and better. So, it's never going to be perfect. It's never going to replace the physician. And I write about that constantly. But they're getting a lot better than where they were a year ago. So it should give a little bit more peace of mind that we're heading in the right direction on the information that's being provided because the models are just training and they're getting better. But yeah, always look at the citations and the sources, click into them. And if there is a specific article that's really caught your attention, go deeper and read it.

Abbie: 28:40
Yeah, that's basically one of the biggest things I do. You know, what I'm taking away from the conversation today is AI is another tool in a patient's toolbox that they can use to improve their outcomes and understand their disease more. But, like any tool, it has to be used effectively and safely. Before we wrap it up, if someone listening has never used AI or maybe they're hesitant to use it because it seems complicated, what advice would you give them? And what's one simple way they can get started using AI to become a stronger advocate for their own healthcare?

Stephen: 29:25
I would suggest just picking one tool, and whether that's ChatGPT or Claude, I would definitely suggest paying for one of the plans. They are not expensive. I think some of them can start as low as $15 or $20 a month for a user. I like the paid plans just because some of the free tiers are limiting how much output you can get, or there's ads and whatnot. It may not be able to record memory on previous asks, so it can't really help you with further context. But I would pick one tool and just stick with that. And the next piece of advice is just try to use it the next time that you would use Google or you have a question that you're unsure on. That's where you would use the tool and ask: You know what, I'm stuck here. Help walk me through this. Literally, have a conversation with it. And you'll be surprised on how unstuck you could get relatively fast just by asking some basic questions on what's in your head.

Abbie: 30:37
I was just gonna say I saw an article that was basically saying that. It was a reporter that was reporting on AI saying that he had asked the chatbot: I know nothing about AI, how do I write a prompt? And then it walked him through exactly what he needed to be doing to ask questions. So it was really interesting.

Stephen: 31:07
Yeah, and I suggest that too. Like here's the example, right? If somebody has a specific disease or a question, we call it prompt the prompt. And it's like, OK, I'm trying to really dig deep and do some deep research, which all these tools have like a separate feature within them to do a really deep dive in a specific subject. But you could say: I'm trying to learn more on XYZ. I'd like this as my outcome. Help me structure a prompt that you build that I would then copy and paste and put it right back into the session, the chat session, and you'd be surprised on how thorough and good it is because it's going to take in a ton of context and build that out for you. You could even ask it before you build that final prompt: Interview me and ask me what I want to see in terms of an output. What's important to me, what do I want to look at? Are there any specific sources, citations I need included? So pick one tool. Stick with it. When you're unsure, prompt the prompt. And it's all about context and really feeding it as much information as you can because garbage in is garbage out.

Abbie: 32:16
Great. You know, Stephen, thank you so much for joining me today. I really appreciate you sharing your knowledge and helping the patients better understand how AI can become a practical tool for them and their caregivers. One of the biggest takeaways from today's conversation is AI isn't about replacing your doctor. It's about helping you become more prepared, more organized and more confident when navigating your healthcare — whether it's preparing for a doctor's appointment, organizing your medication, medical records or just understanding your diagnosis.

If you would like to learn more about today's topic, I encourage you to read Stephen's columns, The Digital Patient, in IG Living magazine.

And, if you're not already an IG Living subscriber, we'd love to have you join our community. You can subscribe to receive the magazine delivered right to your mailbox or choose a digital subscription to read — whatever is most convenient for you. And all of our issues are filled with trusted information, practical advice and stories to help patients and caregivers navigate life with a chronic illness.

Thank you again for joining us for another episode of IG Living Advocate Podcast. If you enjoyed today's conversation, please subscribe to the podcast and leave us a review.

Stephen: 33:39
Thanks for having me.

Abbie: 33:40
Thank you so much, Stephen.

Abbie: Thank you again for joining us today. Additional information regarding this podcast can be found on our website at www.igliving.com. If you have a question that was not answered, please contact me at acornet@igliving.com.

Look for the next IG Living podcast announcement on our website for the opportunity to submit your questions.

IG Living Advocate is a copywrite production of IG Living magazine, published by FFF Enterprises, the only magazine for the immune globulin community comprised of patients who suffer from chronic illness and their caregivers.

Disclaimer: The views and opinions expressed by the guest speaker are their own and do not necessarily reflect the views of IG Living magazine, FFF Enterprises, or its affiliates. (The language I came up with. Feel free to change it)