Latest / Targeted Talks / S7 Ep13: Leveraging AI, Social Media, and Virtual Care in the Information Age

Transcript

S7 Ep13: Leveraging AI, Social Media, and Virtual Care in the Information Age · Targeted Talks

• Juneja Mehta 00:00:05 Speaker 2: welcome to, targeted Oncology's treating together podcast. I'm Dr Pallav Mehta I'm the medical director of reimagined care and a practicing medical oncologist, and I'm excited to have the famous or actually the infamous Dr Sanjay Juneja, oncologist and many other things, but I'll let you introduce yourself. 00:00:28 Speaker 3: Thank you for having me. I'm a medical oncologist and hematologist in Baton Rouge, Louisiana. My wife and I both, live here. She practices full time. I practice part time also with Citrus Oncology. I'm also a vice president of AI and medical informatics at Tempest, and a co-founder of TensorBlack. Basically all initiatives to be able to really try to leverage AI and tech, to be able to just streamline care and democratize it and, and help some of the pain points that we have in practice 00:00:54 Speaker 2: I feel like there's so many places we can go, with our conversation today, but, one of the big areas that I know oncologists think about that I particularly think about, I have a few different roles in my clinical space. I'm a breast medical oncologist, but I also run and started the integrative center. And so these last few years, the whole idea of information and how patients get it and how to filter what's real and what's not has been top of mind for me. And I know for for many oncologists, I think that today we'll kind of talk about social media and how, given your experience, how social media has impacted that and how we could potentially use it to give the right information to the right people. before we start that, though, I am curious, social media is not a place that oncologists, I don't know if there's that many. And I'm amazed that you have all these followers and what led you to that space? Why did you do it? How did you do it? 00:01:50 Speaker 3: Well, believe it or not, it was my, my chief fellow here in heme-onc that my wife had to move with the kids and start practice kind of away. So I had four days to kind of kill time, so to speak. And I started it just by doing the Gemini side of me. I know that sounds a little like, oh, but it's really because it was theatrical. So like there were some dances and this and that, and that led to like actually a bunch of followers kind of overnight, which made me nervous and uncomfortable because I was brought up, in an Indian household where you have this concept of nazar or evil eye and any kind of attention or things in vain is not is not comfortable. You don't want to take from the world, right? So how do you take something like that and make it something that's like impactful or productive? And so, it just kind of didn't take long to think about making content rather now for myself, which was just the anonymity of it all. Um, was instead kind of directed at helping others that were like already on these platforms. So we talked about, you know, screening and how it really catches cancers, ideally in a curable state. And it's not just to tell you that you have a terminal cancer, you know, busting myths basically was kind of the big, start to it. And then a lot of discomfort around the pandemic about blood clots and vaccines. So as a hematologist, I'm able to talk about, you know, what are called thrombotic disorders and antibodies and all of these things. And I just found that there was a real kind of need or appreciation for things broken down in digestible manners that really led to scary things being a little less scary. And, and I think arguably there's, you know, little more scary than cancer. And so I guess between twenty nineteen to about twenty twenty three, I was just really trying to pump out a lot of information for those common scenarios that people find themselves in, either with cancer or with a loved one that has cancer. 00:03:30 Speaker 2: I'm curious, which I know you're on TikTok. What other, if any, like how did you decide which ones? 00:03:38 Speaker 3: Yeah, TikTok and Instagram, primarily. and believe it or not, my demographics show the age range of the highest viewer is forty to forty five. So you'll find, you know, people of all ages, on the site. And then I have had a podcast now for about four years and I bring in just experts like yourself that are, oncology experts and, but all the pieces also pharmacists and entrepreneurs to show what a collaborative process this is and the complexity of what really the cancer system is and how many, how many different, parts are at play to just design a new drug and be able to deliver it effectively. So, that's on YouTube primarily and all the all the podcast sites. I believe we're in one hundred and ten countries, five million downloads, something crazy. it's just sharing the things that we've kind of uncovered and, and learning about that from people that are really fighting the good fight. 00:04:23 Speaker 2: it is interesting when you, you give information to a single patient, you affect that patient and maybe their family, but you do what you're doing and it's, orders of magnitude, overnight, which is great. 00:04:34 Speaker 3: It's great. It's great. But we also find ourselves in a different time because before the reason I said till twenty three, twenty two, twenty three with the content was information was hard to find and pinned down. And now you have an information overload problem. So the challenge is really ensuring or trying to help the kind of, the truths or the pearls or even the direction to go seek the truths, be a louder voice in a sea of otherwise a lot of information, be it misinformation or correct. It's just how does one navigate that space without the understanding or background? I always kind of joke with my kids. I'm like, if I were to drop, you know, dropped in a forest, I wish I could say I could find the tracks of when something was here, you know, recently, or go find an animal for us to eat. But like, I wouldn't, I just wouldn't know how to navigate that as much as I've kind of been, you know, peripheral to it from Netflix. And it's the same thing with cancer and these scary journeys. How do you navigate that without assistance? So I think we're going to see this kind of evolution of not necessarily just having to put the information out there, but possibly help and assist in obtaining that information, especially in an era of, as you know well, of precision medicine where we're talking about precision in a whole nother way than we used to. before it was precision mostly meant the tumor, you know, molecular features. Now we consider things about the patient and about the pharmacogenomics of how they metabolize these things. What drugs are they on? What are their family history? So you have all this complexity that actually can have therapeutic implications, meaning like it can dictate what what treatments you get or not. And so how would you know in that sea of what is the most specific thing and applicable? And I think that takes a lot of, you know, a lot of navigating in that process that that needs assistance. 00:06:13 Speaker 2: there's an article, I think it was a decade ago in the New York Times about this. I think the term was dilution of expertise. as more information becomes available, I think we're all guilty of it to some extent. I know when I try to change the toilet at my house thinking I could do this, I went on YouTube and it did not go well. I should have called the expert. And you know, but that's a toilet. And not a lot of harm came to me at least. but you know, in healthcare experts, the idea of an expert has evolved a lot. And I think you're really using these platforms to, be an expert in a way that a patient may now look at an expert. So I'm curious how you you think about that evolution. obviously we are. I mean, we'd like to think we are the experts in what we do in oncology, but I'm not so sure. Everyone may see it that way when they're in an office visit. 00:07:04 Speaker 3: Correct. And I think, you know, at the end of the day, it's my opinion that that we need to have the humility that we're going to have inherent biases like intentional or not. You know, one example I give sometimes is with, with this new, newer, at the time, drug that was added for myeloma, like it was a rare chance of toxicity, like of a scary one. but it happened on the first infusion twice, two times in a row on two different patients, despite all the prophylaxis, despite everything in the label. And so obviously I'm a if all other things being equal in the guidelines where it's like they're both first line, they're both effective, you know, achieving a total disease control for duration, blah, blah, blah. How am I not going to have a little bias, right? Like, and so that's just one example of how a million different things can kind of inject biases and preferences and these things, which sometimes it's good, sometimes it's je NE sais quoi. And you just have the sixth sense of knowing what the patient's profile looks like, this and that and the other. But with something like cancer, you really want to distill it down to the most objective means necessary, which is why I think this whole concept of, of expert is evolving, especially with AI, because data, you know, an expert is, is they have a higher volume of data, meaning case examples of patients and the longitudinal or continuous outcome data, meaning like what ended up happening to those patients over time to be able to best, recognize patterns and then therefore have the like most likelihood of success with the next patient. So like pattern recognition, data volume and continuous learning, which is a lot of these things with AI terms. And I think the key now is going to be that it's not just the anecdotal experience in the eyes and experiences of that person in front of you, because the accessibility of the aggregate of all of the other cases and all the other, you know, like evolving, continuous, outcomes are more accessible and experts are able to access it more. And so I think there's a humility that's going to go end to end. But to answer your question, I think the best thing that one could try and be an expert in is to ensure that the understanding of the goal in front of us and the options that we have is met by the patient in their family. And I think it actually helps moral injury, even to a degree, because, you know, we all have guilt, even if we had a double survival of average. But but they passed away. Even if it's twice as long, you still you still somehow, at least me feel at the end. I'm like a little shame. It's like they passed away from cancer even though it was twice as long. So like, how can this collaborative process of just continuous learning, can you effectively deliver the information needed to, one, make someone feel good about the decisions that are being made? Number two, be comfortable. and that's the challenge. And you know, because we can give statistics all day, we can give outcomes all day. But I almost think breaking down the words and the concepts can somehow, again, take away the, the fear and angst that comes with big decisions and scary drugs. 00:10:05 Speaker 2: it's a really good point about the, the patience, you know? So it's sort of an expert, if we're an expert on something and we're waxing poetic about mechanisms and, and lecturing a patient. What are they actually getting from us versus being able to explain it? I will say, I've never met a physician who doesn't think they're amazing at explaining things to docs, and I, and I thought I was one of them. a few years ago, I through some other processes, I kind of tried to actually evaluate that. And I realized, oh, yeah, it wasn't very good at that at all. And I think that, again, using these sort of media, you know, that that goes to the patient is a, great way to get the message across. 00:10:48 Speaker 3: But then keeping it accurate, making sure it hasn't evolved, you know, all of these and it's never going to hit one hundred percent. You know, I have one car t example that people have have, said it was very helpful in understanding, but it was billiards or pool, pool table balls, you know, and I can do car analogies, but I can't do cooking analogies, for example. So I'm going to miss, you know, like the culinary experts that are, that are going through these difficult journeys because of a, of a lack of ability there. So that's where I think we all have our, arrows in the quill that could really pierce. And we just really need to master how we can, at the end of the day, let let the what and the how land and not just the, the outcome. Because, you know, we're learning more and more in just the delicateness of human life and finiteness, that the journey is arguably just as important. And even after, like even the fears, people live in debilitating fear, sometimes not even being explained in that process. What to expect? What are the chances of something coming back? This pain is not the pain that actually led to the diagnosis of your cancer, which is terrifying. So now pain is registered in a different way and an augmented amplified way because it's a Pavlovian evolutionary instinct. there's just so many places we could put attention and hopefully palliate, the process. 00:11:59 Speaker 2: I didn't think about the kind of the difference in, you know, a patient gets diagnosed, the fears that they have, the places they'll look for information, but they're constantly in our cocoon. You know, they're seeing us all the time. They're seeing, us, our nurses, they're there every week, every other week. So there's this opportunity to intervene. But then when they're done, you see them in six months or four months, and now there's this huge void which gets to this issue of credibility. I'm curious what do you tell patients both in the clinic when you see them and then on social media in terms of how they should separate, what's credible? What's the, you know, quote unquote, fake news versus the the truth. Is there certain like guidelines you're using or, you know, practical? 00:12:45 Speaker 3: I would say rather than help them separate, which is a challenge I'm yet to figure out. I do try to make it a point that what you hear may not be fully applicable to you. that to me is a lower is lower than ideal. But if I can meet the threshold of. This is something that for several reasons may or may not apply to you. I mean, we know non-aggressive cancers evolve from pancreas, the pancreas, right? That are not necessarily pancreatic adenocarcinoma, poorly differentiated. but you Google pancreatic and you swear that somebody says to you, well, my doctor said the survival you may have a very, you know, that unusual type, the phenotype and even histopathologically everything is not as, as relational. So, I mean, there's just so many factors and so many we haven't uncovered, I could go on forever, but like autoimmune diseases and the treatments on those, like that's not, that's going to be a factor that's different than someone not on autoimmune treatment. So at the bare minimum, I think the hope is understand that there is, you know, generalizing that occurs that very well could not apply to you. But number two, then over time start identifying what those modifiers or variables are that do make you unique to maybe an average or, you know, basic case, so to speak. I think that's the kind of stacking that's required to be able to help others really get ownership, a grip and ownership of their disease and outcome. That's why I feel like we might fail a little bit. It's like, it's kind of like you have this and we inform you what the status is. And that's hard because it's your body, your life, your diagnosis, your calamity. so you really want to help bridle that everything involved around that emotionally and psychologically, you want them to bridle it, not just be like, you're going to come in six months, I'll bridle it. And then you just walk away with the horse not knowing how to ride it. I'll tell you on social media, I've heard a lot the word abandonment, like, like it's a thing where they're like, you know, I saw my oncologist every week or two or three and now it's like off to the wind. You know, I gave you chemo and like, see you six months a year. And I that pains me. And I'm sure it pains you as a compassionate, oncologist. but that's a whole nother podcast and probably an entire series on just the system itself and how we can hopefully get more numbers and volume 00:14:49 Speaker 2: it's an interesting point that, I'd written a book like ten years ago on this. You know, how do people live life after cancer? And one of the chapters was, I'll see you in six months. And I think it's tough because, you know, there's a lot of patients who come in and see us and, we kind of think of ourselves as most necessary during treatment. And so we're seeing those patients. But it does raise an interesting point about maybe seeing those, maybe after treatment is just as important in figuring out, you know, systemic kind of mechanisms to, help us do that, whether it's with our NPS or, offloading work. But yeah, it's, it's an interesting problem. And I think it's going to be more, more of an issue, more patients, less, less of us. Um, you know. 00:15:31 Speaker 3: we have statistics released recently. the incidence of these common cancers are going up, right? Like by about roughly one percent. And, and just all the common stuff like breast, prostate, pancreas, melanoma, the survivals improving, right. I think the first time over seventy percent by at least the ACS data. seventy percent will be alive five years after, you know, a diagnosis, which is pretty awesome. So the scary cancers, deadly cancers are, are less, less death, but that just means more survivors. And to this exact point and more relapsers and all, you know, all these things. it's a challenge. that's a challenge that keeps me up at night. 00:16:06 Speaker 2: I like what you said about the idea that maybe it's not, there's definitely misinformation out there there's people just touting, cancer cures from this or that thing they're selling, but there's actually an entire area of, online, knowledge that patients get from good websites. But as you said, it doesn't actually apply to them. and I'll bet especially in this era where patients are starting to become more savvy with this stuff. Maybe they are going to the right places because we're telling them, stay away from here, but go to this site. But then that site is still not relevant to their specific even in breast cancer. I just do breast cancer, but you think about, you know, estrogen positive disease versus Her2. I mean, just it's a different game completely treatment algorithm and outcomes. I hadn't thought of it that way. I'm curious what you would say to clinicians then. So what should the oncologist who's in the trenches every day dealing with patients? Like what is their approach to social media? Should they what is their approach both to patients, in terms of the advice that they give and then maybe for themselves, how they consume and maybe even create? 00:17:17 Speaker 3: this was an area of interest, A hot topic, I guess for for, you know, last year and the year before, especially the last two years, I would say prior to let's just say twenty twenty five or mid twenty five. With that said, this will always be the case. A lot of times your patient does trust you, right? Like they trust you. And that is so humbling because like, I think there's little that that should humble us so much then despite or in spite the sea of information and experts and, subject matter oncologists. Despite all of that, at the end of the day, they want your explanation because of the trust that you have. Like, you know, established with them. So for that reason alone, it's kind of crazy that in the current system, for the most part, they have to wait, especially if in survivorship. But, but if they're seeing their doctor once, they have to wait for that fifteen minute window, which we're restricted to, to get all their stuff answered from that person they trust most. And then not another fifteen minutes until thirty, you know, thirty days later. And they could be on treatment, like they could be on a Parp inhibitor or inhibitor, whatever, like oral to where they're just getting their tox checks and they get fifteen minutes. Unless the stuff, the didactic stuff, the academic stuff explaining why we look for, neutropenia or these toxicities with certain medications, all that process when you have put it, whether it's social media or not, when you put it in media or video form, now you have something that's evergreen that can be revisited, right? To your point of everyone thinking they're a great explainer. I think the same too, right? I'm like, I know, I know, at least if I do nothing else. But every so often you have the patient or the family that says, so is there a chance I'm cured of this? And you know, you've had the conversation where you explained, you know, micrometastatic disease and you can't see it, but it just shows that it's like it's not I don't know if it's the US explanation problem. I don't know if it's the ridiculousness that we believe that, without having cancer ourselves, what, what hope can bring and what, you know, edge cases can bring. There's just so many factors. But the point is it seems like there cannot be enough of reinforcement, which there is none of that, right? in most oncology clinics, because we're treating we're active treaters. And then the reinforcement part of the process deserves its own time slot, which I think many oncologists, all of us would take if we could have it. So to that point is if it's something that you are saying all the time, like when I explain myeloma, you know, why, you know, antibodies matter, why am I calling them immune compromised if they don't look sick like a severe neutropenic patient? When I'm explaining all these things all the time, what antibodies are and monoclonality. I'm doing that over and over and over again. I expect them to catch it all in that initial visit, which is ridiculous. Or I could say it and I could have it in a video form, and now they can hear it again and they could share it with that. You know, that that son that's concerned, that comes and wants it again, rightfully so. And we get a little like, don't get upset, but we know the next patient's going to get upset because we're going to stay in that room longer. So like, there's just so many consequences that could be Potentially nullified or addressed with having a kind of this evergreen nature of, again, the explanatory stuff. That way, when you talk about the decisions, the decisions of which medication to use and all these things that are that are specific, you have a better concentration of that in your clinic. 00:20:25 Speaker 2: What a great idea. It's almost like the, it's like a virtual, it's like a smart phrase, like a video smart phrase. Yeah, almost is what you're doing because you have, we use them all the time in text in our EMR, you know, dot whatever. But this is the dot for the video. because you're right, a lot of what we say, we all have our style, we all have our analogies, we all have our, metaphors. And, and I think, you know, we do say them a lot over and over and, and having a. Yeah, you could almost say that, you know, if you're part of a health system or something, you know, have each doc have a video library that, you know, that you could, that your patients could access about you. I'm curious though, how would you, you know, some like you're clearly very comfortable on camera and, you know, engaging and. But what about the ones that aren't. Any tips that you would give to just say, you know, go ahead, just do it or, or it's so. 00:21:20 Speaker 3: Funny. my very first startup that I was involved in my job was to go around. well, at first it was just to get doctors to the production team would go and get expert, you know, physicians from all different institutions speaking. And it was extremely laborious and challenging, even though they're really good in clinician room. it was just such a challenge. And so despite the prep and pre-call and all this stuff, so then I would just actually fly with them and be right behind the camera and just pretend to be patient. I'm like, so tell me like when you're thinking about receptor status, what are you looking at for breast cancer receptors? Like, why does that matter? Why does it matter? If somebody asks you, you're telling me I'm curable. You're telling me I can have surgery to to potentially cure me? But you're saying wait three to six months and use a treatment that isn't one hundred percent sure they're going to work on me. Explain that to me. Right. And all of a sudden, now, now they're able to even, like, look at me and, you know, beside the camera rather than the camera itself. We do it all the time. It's just this different thing. But I'll tell you what's what's. And I don't know if I'm necessarily bullish about this, but what's already being explored. I had a Harvey Castro. He's an E.R. doctor, big AI guy out of out of, I believe, Texas. He said a lot of people are exploring avatars to where with enough learning and recording of a physician of us, machine learning has gotten so good in AI that they're actually like able to assemble avatars of me, which again, I don't know if I how I feel about this, but all those videos we're talking about all that content that at least I can guardrail and say anything you've seen in these videos or explanations, that's all fair play for you to be able to deliver as a video, as an animated avatar, if somebody were to ask that. So if I were to authorize my myeloma explanations, my iron deficiency explanations, and my breast cancer receptor status explanations. I toggle all those to say, yes, you can have understanding of this. Now this avatar will speak and it can actually do it potentially better than I could in front of a camera if I'm uncomfortable. 00:23:13 Speaker 2: Wow. Your digital twin. on a virtual patients, you know, in the future, this concept of digital twin, And maybe it'll take off the twenty pounds since Covid that. 00:23:24 Speaker 3: Right that we just can't seem to shake. But I'll tell you what's crazy to your earlier point of, you know, how patience and this was always a stressor for me. How do patients know in the sea of information outside of what they get from their physician or provider? How do they know what applies to them or not? And so because, you know, anthropic and OpenAI both released, the ability to connect medical records to their large language models. So these large language models will be to read and understand the medical records, understand because our large language model, natural processing, because they are able to know what the what the nodes mean. And you can engage with these large language models to ask questions about like, you know, interactions between my medications or my guideline, you know, these, these align with the most recent guidelines, whatever, this, that and the other. But imagine that patients have these potential, their, you know, agents, for lack of a better term. I know that word is overused, but it's an agent that understands it should, in my opinion, be HIPAA compliant. Manner has access to all of your information so that when you navigate through websites, you can click that little icon at the top and say, is this applicable to my cancer type? And they'll say, well, this is not this necessarily this, you know, breast cancer doesn't apply to you because by your records, your Her2 positive and they're talking about triple negative patients. So now you have an assistant, an agent, or an assistant that allows you to be able to know what you're looking at, whether or not or to what degree it's applicable, to. To what you're trying to learn and discover about yourself. And that's why I keep saying, you know, before twenty five. Yes, all this stuff needs to be out there now. We have way like a ton of information as humans and as physicians. I think the best thing we could do is be expert navigators of this cancer process, this entire process. Keep up on our reading and everything, but help them navigate it with us. And number two, with this, insane abundant amount of, information technology is going to help both us as providers as well as patients know the applicability of the things they read or, and whether or not it applies to, to them. the same thing when you and I are looking at literature, you know, we know that our patients are too. And we skim down, we scroll and we find it on Nccn or whatever the piece is, because we want to look at the sub cohort analysis of the particular patient case we're looking for. Think about all that we have to do. When we do that we go back do they have this myeloma cytogenetic mutation or not? Does this apply? All of that goes away because if you have the agent that already knows that stuff, you click and drag and right click and say, apply to this patient I'm looking for. If it's HIPAA compliant, it'll let you know or a step further, just go ahead and binge read all of these five different things and tell me what applies to the cytogenetics of my patient here. Like, boom, it's going to surface it. So we're talking about we're talking about a rapidity like, like an accelerated manner of achieving the precise metrics that we seek, both as patients and as providers, to take care of our patients. That's going to be so much faster. And then the delivering the delivery of those, that information exchange is going to be easier too. So this whole data age interoperability, as those things evolve, it'll be interesting, you know, just in a wealth of information and accessibility, how to navigate that together. 00:26:37 Speaker 2: Yeah, it's changing so rapidly. I mean, I think I got my GPT account two and a half, maybe two to two and a half years ago. And you think about where it's going now this year, next year. If you were to the patient that comes in that truly is convinced that, you know, I had a patient this week, young woman, a locally advanced inflammatory Her2 positive breast cancer who after a lot of convincing, we did start her on treatment. She responded beautifully because she was a er PR negative, Her2 positive. They all respond beautifully tolerated treatment. Very well, I thought. But about partway into it she said, well, my cancer's gone, I don't feel it anymore. I don't want to continue treatment. And that was the conversation we had. And she stopped treatment and doesn't want surgery, doesn't want to do anything else, which, you know, frightening for everyone, including, her husband in the room. what do you say to someone like that? Whether it's in person or, you know, online and someone who might reach out to you, I'm sure you have people that reach out to you for advice all the time. how do you kind of navigate them what are the, what are the skills that people like us should be using? 00:27:59 Speaker 3: I mean, one, I think it's like validation. I think validation is often underutilized to be able to bridge. Now the, the knowledge that may come with better decision making instead of like, we might just quickly reject or we might say, okay, yeah, but let me tell you this, and therefore it should be derailed. So one thing I saw a lot at the beginning was, and it hurt my feelings. They'd be like, you know, the whole thing's a farce. These cancer doctors want to, you know, I mean, you know, the, the narrative, like make money and everything's gone, you know, for twelve to fourteen months and magically like it appears. So I'm like, you're right. I'm like, that sounds sketchy. How can it how can you have ten different spots? And I do a Pet scan and it's all gone. And, and all of a sudden something pops back, even despite you staying on the therapy. But then that's when we talk about, okay, let's talk about what, say an average Cat scan looks at, right. So it may need two hundred million cells in a collection in an aggregate to be able to be large enough in millimeters, you know, for five millimeters on a on a CT scan on an average CT scan to be seen. So that means the fifty thousand one hundred thousand five, all of that number below two hundred million cannot be seen on this Cat scan. And then I'm like, and then, and I'll say to grow to be big enough to be able to see. And they all have something called a doubling time. Yours is not that aggressive. You might be a hormone positive. So like you don't double that number from fifty to one hundred and do the math all the way to two hundred thousand. Now all of a sudden like, okay, I see why I could appear disappeared for a long time because death to cells occurs much faster than the doubling time of growing. And of course, I'm saying all this fast right now, but like, but then all of a sudden, they get like, look. And even then with the Pet scan, it doesn't light up because you have to have enough of this fuel burning, this coal burning with, with FDG uptake of the Pet scan that that's why we don't really think about it in this cancer, say for one centimetre, because even if it's five, it doesn't light up, doesn't mean it's not cancer. It means like it's not hot enough for to appreciate because of the energy burn. So again, these are the concepts that I think like that's, that is where I think the expertise, you know, we could really sharpen our, our blade, but it's hard to compete with, with AI now because I could for the same deficit where I can't speak in cooking analogies, I could go to an LLM and say, using culinary examples and metaphors only talk to me as if I'm a non-medical person. Let's just say at a tenth grade reading level and explain to me what are the top ten reasons I would not get brightness on a Pet scan, even though like I might have cancer and do the same for CT or Cat scan. Now everything I just said is going to be. And you can just iterate and iterate and iterate and say, give it to me another way, another way, another way. Give me an example of how much that doubling time means. Oh, that's why it takes sixteen months, because the doubling time of the now all of a sudden things are starting to make sense. And I did it by validating the the fact there is no cancer. But that has an asterisk because there's no cancer. I can see understand, you know, it's all game for under five hundred million, right? Or if it's an ultrasound thing, it's even, you know, more so like, and then that's where that's where this may come. Then the hardest thing I think that we have that is unfair to patients, but there's no other way that I can think of to do it is the statistics, which means of all those patients, you know, of ninety percent, you know, like had complete response, can't see anything. Look how many if you don't do the surgery, you know, progress and then have disease, you know, somewhere else. So I don't know, I might plug that into the LM after this call. Let's just see how, how does it do it? How does it explain, you know, we have to take statistics in our boards. Like that's not everybody, right. 00:31:29 Speaker 2: Right. I see this every, at this point, probably fifty percent of patients ask questions like this and, and our ability to respond. I like the validation idea, you know, of kind of making sure that even we have some awareness of it. understanding of the kinds of things they're asking because I think if we don't know, we're sort of very defensive and just, you know, are immediately willing to kind of push it aside. But if we have a little bit of understanding and you read about it and I think we can use AI hopefully the right way, but it makes me think to even have like a, a list of like prompts, like, you know, as you said, explain it to me like I'm a blank. Yeah. You know, because I know when I talk to AI, I'll sometimes say, explain to me like I'm a medical oncologist, medical oncologist, you know? Right? Versus I'm trying to for your child, you may want to say, explain to me like I'm a ten year old. So that's, that's actually good advice for, for our patients to, to maybe just say, listen, here's where I am, here's my background, here's my job. This is what I do for a living. Now explain it to me. Yeah. That's a yeah, it's a good yeah, it's a good idea. 00:32:31 Speaker 3: I have like seven different project folders, like one for my kids when I'm driving them. That is already set at default for like ten year old use metaphors lyrics. You know, these fun, imaginative examples. My baseline is me as an oncologist knowing tech stuff. So yeah, definitely want to personalize everything. I'll tell you one thing that's interesting though, that that came to me as we were talking about how we can help patients the most in a world full of information. I think one thing that will not be replaced by AI and tech, that the token that we can use as physicians is emotion, like giving someone an idea of the emotional consequences or even arguably the word suffering. Like what is the suffering up front to tolerate something? What is the suffering with a bad decision. That's something that I think is very distant from anything that you could read on statistics. And I'll give you an example. It's when sometimes we get into this discussion of do we want to, you know, technically we're allowed to break, do we want to watch and wait? Do we want to do a maintenance? Do we want to do adjuvant, meaning give you chemo to decrease the chance of something coming back, knowing that it could never come back? Either way, like, but we just don't know which category you are. These difficult conversations. What I then ask is bear with me. I know it sounds silly, but if we stop or don't do anything and this cancer comes back, God forbid, versus if we continue and you have a bad side effect that requires a scary, admission to the hospital, which is the one that we as individuals and potentially as a family will say, I knew I should have like, is it the one that came back? Do you and your spouse and your kids have to live with saying, I knew I should have just tried it, the chemotherapy to reduce my chance. Is that going to happen or are you going to say, I knew it was the right decision? I'm like, that's all. I'm like, there's no right answer. But if you can. And nine out of ten times like I know it would have been the right decision or it would have been, oh my gosh, if it came back, I would regret it for the rest of my life. And my spouse was like, I'll never forgive myself. So it's like, boom. Now I put something through the lens of emotion and quantified suffering, not quantified qualitatively, but in different capacities. Right? And I think that's a humongous, I even use it on myself. I'll be like, when I'm making treatment decisions, which one am I going to regret? Or saying, I had a gut feeling I should have X, Y, Z, you know, and, and when you think of it in this other way that I think, you know, LLMs and AI and information is far away from, it helps with the decision making that at the end of the day, are still just decisions. And that's, that's, I think, something unique to us. 00:35:04 Speaker 2: Yeah. It's the error of commission versus omission. What are we more willing to in a specific scenario willing to tolerate. so in the last few minutes, I knew this was going to go long because we have so much to talk about. but the virtual care model, this is something that I've been interested in with reimagining care the last few years. And, and there's so many places, virtual care, we know there has to be virtual care in some capacity in, healthcare in general, but, the ability of virtual care to deliver information, telehealth, we know that there was an expiration of some of the telehealth regulations, the Covid regulations. I'm curious what you see from that perspective, maybe even just from a burden of the oncologist schedule standpoint or maybe convenience of the patient standpoint. You know, where do you see, you know, the virtual care model and the information delivery ecosystem kind of aligning? 00:35:59 Speaker 3: Yeah. I mean, I'll go one step bigger than that and just talk about the stuff outside the office, right? Like let's what we've said this entire podcast is, number one, there is a paucity of opportunity to be able to, even share what you want to hear, what you want to learn and what you feel is important enough to share or not share to balance the other part of it. Hear what the doctor has to say. Ask the questions about whatever is this important or not. So you have to. It's so unfair. The patient family has to juggle that in a fifteen minute, window. Can you make that window larger to where maybe understanding what something to complain like not complain about, but something to report in the way of side effects and stuff. is known. That is number one, where virtual care, you know, can help where I think even potentially wearable devices will help. How can you track, how can you inform somebody of the things that need to be reported and take priority? And that's what like these, this, this virtual care is allowed. It gives you more opportunities to be able to share things of concern that then permits the visit to be with potentially education, decision making and learning. But we know how many people and this one hurts almost more. Don't get an amazing therapy that's working potentially life saving because of a toxicity. And and at least with the newer drugs, these toxicities happen over time. It's not like black and white, like a reaction. It's like there's grade one, grade two, and it's kind of missed and it's subtle. How do you make sure that you catch that early if you know, it could, it could spare someone having to table a good therapy forever. So that's number one. That's where wearables and virtual care to me are, incredible because they take that two to three to four week window between visits and make a utility of that entire duration. Whereas if you don't have wearables or virtual care, now, you're just at those two fifteen minute visits, three to four weeks apart, right? And they're supposed to somehow surmise all of these different things in that in that three to four week window of time and on their own. Not being an oncologist. Know what is worth bringing up, not bringing up to make sure that they maximize the use of those fifteen minutes. That's insane to me, right? So that's number one. And then number two, I think what it offers is, is, well, it's really it's one and the same. You get to understand your threshold for what you share goes down rather than has to stay high. That's what virtual care allows. What wearables and virtual care, good virtual care companies at least like allows, is also giving that information that I want to be able to say every visit. Tell me about a fever. Tell me about a rash. How many times do you have diarrhea? All these different things that that, you know, my triage team is phenomenal. So like we're able to field, but it is laborous. Like we have to keep a lower census just to make sure anyone can call in during the day and when it's when it's not in the day. My team's not there. I'm not there unless I'm on call. Like, what about those hours? So this is where I think we just get an expansion of opportunity and where I think AI and tech really fit. even outside of that, just, just accepting the concept or need rather of care and interaction outside of those point visits. I mean, that's, that's crazy. Cancer care. Cancer treatment is continuous. Living with cancer treatment is continuous. fighting cancer is continuous. But right now without these things, it's a very just isolated, point to point interaction. 00:39:19 Speaker 2: Mhm. Yeah. You're right. Expecting patients to kind of remember through the chemo brain exactly what happened on Tuesday, Wednesday, Thursday and Friday and then report to us. And I mean, that's you're absolutely right. They have so much. 00:39:31 Speaker 3: How many times have you told someone at first when you started like, why didn't you tell me that? Like, oh, like, I wish you would have told me that. Right? And it's like, how you know, then you catch yourself because you're like, how rude. Like, it's like, God knows what is on their plate, right? But it's like you have you seen, notice this crazy wound or this like this fungal thing and like, or, you know, the cough, the cough, like two weeks. Well, last week you said it's just unfair. But but but some kind of continuous monitoring, which is why I keep looping those together, whether it's wearables one day, whether it's continuous monitoring devices, whether it's virtual care, whether it's agents that are just like sitting there and basically being a journalist, that there needs to be something between visits that because the process and the complexity of these newer drugs are a continuous process, and these things can pop up at any time. And it's ridiculous. It is it is rude to ask just in any average person, including myself, being an average person, to somehow make balance of all this on their own between visits. I mean, that's just offensive. 00:40:28 Speaker 2: Yeah. And these treatments are getting more and more complicated too. Well, this is great. I mean, again, we could talk for a while. Is there is there anything, a rule, anything you would say to a patient that you haven't said already that something you would you would tell them now around the idea of information, social media, that I could share with the patient, you know, or that a patient would share with me. Um, you know, next week. 00:40:52 Speaker 4: I mean, I've, you know, I've. 00:40:54 Speaker 3: Only, I guess only practiced six years or already six years, however you want to phrase it. But, but I think if there's one thing that I try to share that I want patients to know, and even honestly, like physicians, it's the power of a gut feeling. Like, I just think at the end of the day, the decisions we make in our lives as providers, as patients, as family members of loved ones with cancer, you're in a weird way, you're hedging against yourself, like you don't want future pain and suffering and regret and hurt and ache. And I think the way you hedge against that is any gut feeling to address those gut feelings, be it a gut feeling that I know I need to ask about this, I need, I know I need to tell them about this, I know I saw a trial, I can't be shy about asking because I don't want to offend them. But whatever your gut feeling is, hedge against the possibility that you say, I knew I should have. Why didn't I like? I think I really think that is the compass that could at least protect our future selves from otherwise just bad outcomes or bad luck or whatever else. Like bad things happen and we're all going to experience them. And but in an ideal world, bad things just happen rather than a bad thing happen because I didn't act or I knew I should have or I could have prevented this. And I think if I think if we just remind ourselves that with decisions that we make on incorporating AI tools or tech tools or exploring a virtual care model, even though it seems laborious today, what are the things I can do to hedge against complication and, you know, a, a, a distraught future self? And I think that, you know, when you think about that future self, I think it it protects you from that. 00:42:33 Speaker 2: Yeah. Trust your intuition so as not to escalate issues when you could have trusted. Yeah. And The right way. 00:42:42 Speaker 3: Like if you believe in that snake. 00:42:43 Speaker 4: Oil, right? 00:42:43 Speaker 3: I don't want you the whole time to have not said anything. You know, and and your spouse said, don't. It'll be, you know, it'll offend him. And then the cancer comes back in a year and a half. And now you're saying to yourself, still, I should have asked about this because it would have come back anyway, right? And then like, you know, if that's the case. So like, my point is, these are the kind of things that seem so small. 00:43:02 Speaker 3: Just do it. I don't think any of us as humans, you know, would ever be offended enough to make you feel better about the decisions that you've made in your life today. Like I would jump in here too. I always think my patients for that.