AI Is Not Magic | Dr. Douglas Manuel

AI in healthcare is often talked about as something magical. Dr. Douglas Manuel, a family and public health physician, senior scientist at the Ottawa Hospital Research Institute, and organizer of Ottawa's recent symposium on AI in health research, pushes back on that. AI is a tool, he argues, an instrument like the thermometer or the blood pressure cuff. A revolution, yes, but medicine has been absorbing new instruments for two thousand years, and like all of them, this one has to be calibrated.

The conversation covers why AI scribes have been adopted faster than anything Manuel has seen in 35 years of medicine, why the messy data underneath the algorithms is the real bottleneck, and what leaders should watch for when choosing between AI products. He also draws on his pandemic experience as one of Ottawa's most visible medical voices, and lands on the warning easiest to overlook in the rush to adopt: almost every new technology widens inequities before it narrows them, and the question is never just whether a tool works, but who it works for.

Transcript

0:00

Jamie Brehaut Welcome to MythInformed Science, a podcast about misinformation, disinformation, and the science of what to do about it. Conversations with experts in the study of misinformation.

Okay, so Justin, you and I teach a graduate course in Ottawa, in the School of Epidemiology and Public Health. We call it Epidemics and Infodemics: 0:02

Influencing Public Health Beliefs, Decisions, and Behaviors. It's a long title, but you know, it's a course that we worked on for a long time. And we're teaching graduate students, we're teaching medical students, we're teaching master's, PhD, postdoc, people who are going to be working in the healthcare systems. And the goal, I guess, of the course is to help these people, these healthcare system actors, know a little bit about how to guide people's beliefs and behaviors during a health crisis. How to teach people how to distinguish between good information and bad information, where to get good information and where they need to be suspicious or at least skeptical about the information, that kind of thing. There's lots of things we touch on in this course, but there's one particular class that I know I really enjoy, which is the one you teach, and I think you call it AI, Trust, and Information on the Internet. But you find that one a particular challenge to teach, right? Justin Presseau Yeah, I wouldn't say it's a challenge to teach. I quite enjoy teaching it, but it's the preparing for it. And the reason that's a challenge is that it's clearly just such a fast-moving space, right? So when we first taught the course in 2021, it was sort of still within the pandemic and we started the course itself in response to trying to be helpful in preparing for, or responding to, and preparing for the next pandemic. And at that time, this was before ChatGPT was a big deal. 2021, our first time, and then the next time we did it, it started to become more prevalent. And then this past year that we've done it, it's again, the amount of change between each time we run it is just remarkable. And so the challenge for me is just to keep up with the times, which is also fun to just see how things have moved on so dramatically every time that we teach this. And— Jamie Brehaut And Annette O'Connor used to call that the information has the shelf life of a banana. Justin Presseau Yes, which is about right. I might use that. That is absolutely how I feel when I prep for this. And I mean, an example of that is that to me, the most striking example of that is it was, it is clear to me that when we first started the course, I wasn't even thinking about how AI could be used for good or bad. AI is now a tool that anyone can use to produce all sorts of content, whether text or video or image and everything in between that could be leveraged in all sorts of different ways, to the point where the most recent iteration of our course, we started to talk about this idea of AI agents and workflows that things can get automated without humans in the loop at all. And that sort of creation of AI slop, which became a big conversation in our course, in that particular class, was, you know, we didn't talk about that until this year. And so, yeah, it's a challenge to prepare for, but it just feels utterly relevant, increasingly relevant. And you know, we start that one, I think, is that the one we start off with asking our students to wear tin foil hats because this is a space where it quickly gets into that realm given how quickly things move? Jamie Brehaut Yeah, no kidding. So coming up, we've got someone who's doing a lot of thinking about these kinds of issues recently. Dr. Doug Manuel is a family physician and a Tier 1 Clinical Research Chair in the Department of Family Medicine at the University of Ottawa. He's a senior scientist at the Ottawa Hospital Research Institute, and a distinguished professor at the University of Ottawa, and a senior core scientist at the Institute for Clinical Evaluative Sciences in Ottawa. His research covers a lot of ground, but it involves exploring how big data can be used to most effectively improve the health of communities and reduce inequities. He's recently organized a symposium in Ottawa titled Ottawa Symposium for AI in Health Research, where he brought together a wide range of researchers leading the charge of research in AI and healthcare. Welcome, Doug. Welcome to the podcast. Douglas Manuel Great to be here. Justin Presseau Yeah, welcome, Doug. So not that we necessarily want to remind ourselves too much of the pandemic, but if we bring ourselves back to that point in time and thinking about your role during the pandemic, you were doing a lot of media interviews, had a really public voice around all sorts of different COVID-related topics. And so, we thought we might kick things off just to hear about your experience in being positioned in that way and being that public voice. How did that come about, and what was that experience like for you? Douglas Manuel I can't really remember how it came about. I am a public health physician and a family doctor. And what I, how I practice public health, I think of the community as your patient. And so, instead of saying to an individual, like I would, if you came in the clinic, Justin, I might say, "Hey, do you smoke? How healthy do you feel?" As a public health physician, I would say, use the surveys to say how many people, what the proportion of people in Ottawa that smoke. And so, you're thinking of the whole community as a one sort of patient. And how do you talk to patients? I think, you know, one of the main ways that we talk with patients is publicly. And so, I just really assumed the role as one of the, you know, one of the doctors for the community and started receiving requests. Then I was on the science, Ontario Science Table as well, which got a lot of attention publicly as well. Justin Presseau Can I interject there? Just for those that might be not familiar with the Science Table, can you say just a quick word about what that was, in particular? Douglas Manuel Well, in Ontario, it was supposed to be an independent science review or advisory body, I don't think of it as, but a science review of the science feeding policy. So, it was, you know, that the scientific director and the chair would report to Cabinet or they would speak to Cabinet. And so, it was, it was about 20 or 30 people, sometimes larger, but reviewing evidence that would inform policy. Jamie Brehaut Yeah, I mean, and just really interested in, yeah, that particular voice and experience and speaking to the community as your patient. I mean, do you feel like a lot of physicians take on that role in the same way you do or as effectively as you do? It's a perspective I haven't heard quite voiced that way, and it's a really interesting one. Douglas Manuel Yeah, well, ultimately, I've been practicing for a while, and I think when we, we can talk about my approach, but I think you develop your own personal approach to it. When we're trained in residency in public health, we do have training in health communication. And when I was in residency, too, we would have things like mock interviews or we would be trained with media folks. And when you're examined, the Royal College examination, there's always a role-play role where you're in the media or something like that. So, I think as a specialty, it's definitely one of what we call core competency. But you know, I don't know. I think there's a lot of kind of people that became informal spokespersons during the pandemic as people that were willing to respond. Sometimes there would be a lot of people that wouldn't want to be on the news. It's actually quite difficult. It's quite challenging. As well, you were getting criticized quite heavily. So, I remember one time I was talking to one of the reporters, they said, "Can you find any woman for us or people other than like a white man?" And I, my colleagues and people were not willing to expose themselves, and I felt I had, so I felt also I had more like training, agency, ability, and also just the imperative that this was a really big thing. I'm a public health physician. This is go time. So, that I needed to step up. I think initially, too, for a lot of it, I think for a lot of scientists when we're in the news, you're coached like just talk about your study, don't talk about anything else, just your study. And you're on the news, and first I was on the news for say projections, but then they're asking me everything about the pandemic. And you feel a little bit uncomfortable with that, but I think as a public health physician and over time, you realize that you actually have knowledge and expertise that other people don't have. And during the pandemic, I was on a lot of advisory committees working night and day on it. And even though I was being asked maybe some questions that were outside my field, I have to remind myself that I have more information than the listeners do. And so, just try, I would just try to answer the question that was proposed to me with what I knew, and if I didn't know it, I would say that as well. But you're covering off a huge range of topics just on the fly, and you have to figure that out. Justin Presseau And I mean, you sort of went above and beyond the call of duty during the pandemic and became certainly a locally trusted voice in the media. And I'm sure that itself was challenging to incorporate alongside everything else that you were doing. But have you found that now that it isn't really go time anymore, are you still, do you still have that sort of public presence? Are you being asked to have these sort of interviews and conversations because, I mean, you established that kind of voice, that trusted voice in community, but was it a pandemic-specific thing or do you still see this as part of what you do? Douglas Manuel Well, you know, I've been in the media on and off for decades. And my research is public health research, a lot of it, so it is public-facing. So, every, every, you know, I'd be on the news, I don't know, maybe five or ten times a year normally, and maybe a bit more than the average scientist. But during the pandemic, I lost count at 300 interviews. And you have to remember, too, those are 300 interviews. I would spend a lot of time with the reporters, treating the reporters like they had a lot of questions. So, there was a lot of back, back research and studies and connecting people, and trying to be really available because, and treating the reporters as colleagues. They're the experts in health communication, they're the experts in communication. Jamie Brehaut Fascinating stuff, Doug. I mean, I'd like to take a course from you on media communication and health communication because the experiences I've had speaking to the media have been very different. And partly, I think it's from my 'I want to talk about the specific research and I don't want to go beyond and talk about things where I don't feel I'm an expert in,' and I've been pushed into situations where I'm just talking out my ass about stuff that I didn't want to do. And yeah, so speaking to somebody like you who's had all this experience would be really, really interesting, and I'd learn a lot, I bet. Douglas Manuel Yeah, I try to go on every show. Like, you know, I would be on a lot of talk radio. Talk radio is actually, talk radio is a lot of fun sometimes. Talk radio has advantage, too, is that I think as scientists we sometimes really are concerned or really the formatting of things are very important. And if your study is not taken the way that you intended it, then that's difficult for us. And you can't tell what the clip's going to be. On talk radio, you can. It's live and you're just talking. So, in some ways it's, in some ways it's actually easier in some ways because you can talk, you know, talk what you want and the listeners are going to hear it. But, but there's quite a range, like talk radio is different than CBC News or something like that.

Like news radio, like top of the hour, here's the, here's the Blue Jays and after that we got Doug Manuel and you're cut. So, it's a different, quite a different sort of audience and format, but I think very valuable. And you know, again, that's an audience that needed to hear things as well, so I was very happy to go on those. I would often go on just for a few minutes in the morning. I'd be at like before 8: 0:53

00 and go on for a short talk. They would know me so well they would just call up quickly and say, "Can you jump on?" Jamie Brehaut Oh, fantastic. So, so I do want to get to the topic a little bit more about AI as an important way forward in healthcare. It's certainly been, we're hearing lots about AI in the public sphere, but in the healthcare sphere particularly, it seems like there's lots of apps being developed, lots of applications for AI being developed in various little silos and across different disciplines within healthcare. And we were mentioning at the start about the meeting you organized, the symposium for AI and health research. Can you tell us a little bit about how that meeting went, why you wanted to organize it, those kinds of things? Douglas Manuel Yeah, I think, well, one, I mean, a shout-out to the sponsors, I lead what's called the TDAI Hub at the Ottawa Hospital, and it's funded by TD Bank, and I just wanted to acknowledge we would have never had the meeting as well without them. But, but part of that funding was to have a meeting and public event when we applied for the funding and we said we'd do it. That was well before AI really started kicking up. But we felt, I think this fall especially, there's just a real appetite from everyone, scientists for sure, but I think people. So, there's, so we thought coming, we thought that I really feel strongly that this is huge, and this is affecting all of us, and we need to have a conversation about it from healthcare and from the perspective of healthcare healthcare research. We need to talk together again as a community and it's a shared issue. So, really it was getting to know each other, and there's so many people getting involved with it, to connect scientists together and connect scientists with the public. So, you know, we thought, I literally thought like I had this really big advisory group because I figured like if they're on the advisory group, at least they're going to come and maybe we'll have 10 or 20 people. But then we put out, we weren't sure, so we put out a request of interest, and we received over 400 people, and by that time we booked a space for 100. And it's hard to get people out sometimes to these meetings, and this was nothing like I've ever seen. So, so we actually had to instead then be selective. So, just incredible. And when I talk with anyone, what I'm listening to them, you know, in our house, we have a term that my daughter brought home from her kindergarten teacher, Madame LeBrosse, and it's 'excurvis' - excited, nervous, and scared. And when I'm listening to people, I hear it, like the scientists, everyone. And there's always like there's some, you know, I add it up to 100 and I think for me, it's maybe 50 or 60% excited, I'm quite, I have bias towards optimism, and then maybe 30% nervous, and maybe 10 or 20% just downright scared. And so, so you want to talk when you're scared and nervous and excited, you want to talk. And so, we had, I don't know if this is going on a bit, but at the very beginning of it, it was at the Museum of Nature, beautiful space above where the Parliament was when the Parliament burned down. And I tried to do a pep talk because scientists are all introverts and no one likes to talk and things like that. So, I said, look, I've been waiting, I've been trying to prep myself for this for a week or two knowing that I need to get out and talk with people, and my goal is to meet people, talk with at least 10 people that I haven't met before. And you're all introverts, too. So, I'm just asking you to kind of like suck it up and get out there and at the end of the day, we'll all have a drink. And we'll all be tired, we'll all have a drink. And I think really people were probably primed as well, but were willing participants on that. So, it was just a really noisy room for most of the day. I've never seen anything like it. And I think that's just a testament to where we are right now. Justin Presseau Amazing. Yeah, excurvis, I feel like if you allow it, that feels like the title for this, this episode of the pod, the podcast. Douglas Manuel Just give credit to Madame LeBrosse. Justin Presseau Okay, we'll do. Yeah, and having been a part of that event, I can reflect as a participant that yeah, the energy in that room was just palpable. And the people that you brought together across different disciplines was just yeah, a really exciting space and and highlighting just how yeah, excited people were about the opportunity to connect and to talk about AI. And I guess with that in mind, as you reflect on the day and your work generally, what are, for our audience, what are some of the kind of key use cases, key applications of AI at the moment that you see are are transforming at the moment or poised to transform healthcare and health research? Douglas Manuel Yeah, you know, it's it's here, like it's being used more than you realize already. That's one thing, but we, our public event the night before, we talked with the panelists, we talked together and we decided to frame AI scribes as a key discussion point. And an AI scribe is like, so for 10% of the visits in Ontario right now, mostly family docs, when you go in, the iPhone or it's going to be sitting there, and your doctor's going to say, "Do you mind if we record this so so the AI can make summary of this?" And that's what an AI scribe is. It's listening to the conversation and then it's it's all kind of privacy, like so Ontario has approved 20 AI scribe applications, and a lot of approval was like privacy and security and things like that. And so, it's nothing's kept. It's just kept for the time of the recording and then it's summarized, you can go into the charts. So, you can have as a patient you can have a summary and the doc can have a summary as well. And it has been explosive. I've never seen anything like, I think I've been practicing for 35 years, training for longer, and I've never seen an intervention come in so fast, so quickly. Jamie Brehaut Yeah, my family doctor just kicked it in just a couple months ago. Justin Presseau Yep, me too. Douglas Manuel Yeah, yeah, so it's projected 89% in 2 years. So, it will be, it will probably be every visit. And that's not like every visit for like, you know, your, your endocrinologist for your diabetes, that's every visit. Like we don't see things like this. And there's a lot of concerns, but most of the, the assessments so far are positive. Patients like it, they can, the docs actually talking to them and they're not looking at sideways at the screen. We've all experienced that. Doctors do not like doing that as well. We like talking and looking at people. And then so patients actually are reporting pretty high and saving, saving time for the doctors and for family health family practitioner, your evenings are doing charts. And people are saying that they're getting because it's charting for you. And people are saying that they're getting back some time and most of the time when I talk with someone about it, after they start using it, they say, "If you take this away from me, I'm going to quit." You know, it's just really positive. Now, there's a lot of caveats, but net incredible. And I can't even imagine how that's going to play out in different ways, just that one application is going to change what medicine looks and feels like. Jamie Brehaut So, so, yeah, AI scribes for sure, it's, yeah, amazing implementation has just been adopted wholesale, it seems, and it's amazing to watch. There's lots of other sort of applications of AI, I think, that are not getting the same kind of immediate adoption across the board, and I wonder if you're hearing or if you've heard from through the symposium or anything, if there's any sort of key sticking points or any particular barriers to new AI applications being adopted in ways that would, could potentially really improve other aspects of care. Douglas Manuel Well, yeah, I think there's a couple dimensions there. You two, what you both do, I think is one of the big dimensions in implementation is always really challenging to do well. You can never do it as well as you want. And then, I say virtually every new technology that comes in increases inequities. At the beginning, because there's access, the people that are going to use AI and access it are going to be the ones that have more agency. And you know, AI scribe, what happens if English isn't your first language? So, so, um, it's really difficult, the implementation to do it well. I'm a data guy, and I, the, it's actually pretty pretty complicated to integrate these things, and our data isn't where we want it to be, and our systems aren't where we want want it to be. And you can, you know, well, one of the things that we're doing with the AI Hub is like that the just the technical part to bring a new algorithm in, I say usually takes us 9 months a year, and I've been trying to compress it down, I say it should be 9 seconds, but no, it's still like 9 months a year. And it's a lot of the again the processes, the the people that you're working with, but the technology can be difficult as well. Jamie Brehaut It's interesting. I mean, I think a lot of people think that, you know, we've heard that there are electronic medical records that all the data is being saved and that it's it's there, it's just a matter of plugging it into an AI and suddenly everything everything's solved. But the data issues specifically are really very complicated and the variation with which data is coded across sites and across jurisdictions and provinces and all these things are real challenge, right? Douglas Manuel Yeah, like Ottawa Hospital finally joined the Gemini system, which is 30 hospitals in Ontario, it's going it's spanning across Canada to combine their data. This is a field that I've been working for a long time, but we had Fahad and Amol who lead it come up, and one of the slides was like Fahad was like, "Here's the 30 ways that sodium is recorded in the charts in Ontario." Right? And then and as data scientists, you know, the adages 80 to 90% of your time is data cleaning. And and as there's more data, as the data becomes more complex, the problem can become more complex as well. AI could help, I think I'm optimistic that AI can help with that data cleaning process as well and our ability to generate code and generate infrastructure is also increasing exponentially more powerful and quickly at the same time. So, so, um, you know, I think I'm kind of rambling on, but it's a pretty it's a there's a lot of there's a lot of things to there's a lot of things to you're experiencing the whole excursus if this is the theme of the theme of the show. There's lots of things to be optimistic about, but there's also there's also ones that that the challenges or the the concerns are sometimes the same scale, yeah. Jamie Brehaut Yeah, no. What do you think about, so I don't want to lean into the fear part right at the right at the end of the interview, but I mean, what do you think of the idea that AI has the potential to contribute to the overall flood of information, just the complete swamping of people in a way that we see in social media and that sort of thing? Is is there a danger of that in the sort of AI application to healthcare? Douglas Manuel Well, I, how I want to answer that is, you know, where I think we are go time right now, I think go time right now is child mental health and child well-being. I think it's the largest health largest issue in the world globally right now, more than global warming, more than war in Ukraine. And I'm concerned that our children have been taking hits for the last 10 years in the way that society has been changing. I think screen time and social media gets slammed a lot, but I think it's kind of emblematic of some of the challenges that kids are facing. As a public health physician, if we have 1% increase in heart disease, 1% increase in depression, that's a tsunami. We see, you know, that's millions of people. And children it's doubled. It's gone up like from the teens to like mid-20s. So, it's like multiple tsunamis. And this is our children. This is our future. And and I'm concerned that that there's a lot of signals that AI's going to make that worse. Will they have jobs? Will they how can get an education? Are they going to be on the screens even more? And even if we're concerned about this, it seems to me like it's go time. Because it's just the magnitude of it is just unbelievable. Jamie Brehaut Yeah, for sure. One other issue that I thought maybe we could touch on was just this notion of the current marketplace of AI applications. My my sense is that there is increase there are increasing players in the field that are designing AI applications for every aspect of healthcare in one form or another, and they might be AI light or they might be AI heavy in term in terms of the actual I don't know, level of sophistication of the AI applications, I guess. But for organizations, hospitals, healthcare organizations of one sort or another trying to choose the applications, the AI applications that are best suited to them, that are going to work, that are going to save them money, going to save them time, push them forward into the next to the more efficient and better care for patients. What would you say to the people in the organizations that are faced with choosing one AI product versus another or choosing amongst all the different potential options that are either there or going to be there very soon? What are some of the things that people should be watching out for? What are the things that maybe oversold at this level in terms of AI products, those kinds of things? Douglas Manuel Yeah, um, my first thought bubble is good luck. The no, I I would say that you follow your game. There's one thing I say is AI is not magical really for the most for the most part and and it's a tool, and it's an instrument, and we've been bringing these in for about 2,000 years, and we have a process for that. And don't get, I think one part I would say is I've been saying I'm scared about a lot of things, but I would say try not to treat it like a mystical thing. This is huge, but it still, you know, I think it is a revolution, but it builds on still our practice of medicine, Boerhaave, and physiology bringing in a thermometer. I talk about thermometer a lot when I teach AI, and um like I say every medical tool that you have, blood pressure cuffs, thermometer, bathroom scale, it all needs to be calibrated. It all needs to have regular assessment for calibration. And and it some of those are easy. You step off your scale, it says zero. You step on it, and if I step on it and it's 100 kilos more than yesterday, it's a calibration problem. It's very difficult to do that with AI, but the principles are the same. Jamie Brehaut We have the, we have the skills to implement AI usefully and effectively. We just have to use the skills that we already have. Douglas Manuel I think that's easy to say. Like I said, like I I did not see scribes coming. Like we've been working on scribes for that the concept of transcription for a decade or two, and then all of a sudden boom. So, you know, it's every month or two, probably every 2 months, it's oh, I didn't see that coming, and it just changes things. So, when I listen to the the CEOs of the companies of Microsoft or OpenAI or others, and when I hear them saying we don't know where we're going to be 6 months or a year from now and maybe we're going to have you know, quote unquote, super intelligence, like everyone's predicting super intelligence in those communities within years. And and my own experience, I did not see things coming, and when you hear them saying that we don't really know where we're going to be 6 months or a year out or year two out, I think that's noteworthy. Jamie Brehaut More reasons for excurvisness. Douglas Manuel Yeah. Jamie Brehaut Look, Doug, we could talk about this stuff all day. It's fascinating. Really, really appreciate your time. I think this has been a really helpful discussion and gives me lots to think about in terms of where we're going and how we manage this revolution that we're in the middle of. Yeah. So, thank you very much. Douglas Manuel My pleasure. I mean, this is I think one of the main things we need to do right now, talk. Jamie Brehaut Yeah. Yeah, talk. Douglas Manuel Yeah. Jamie Brehaut Yeah, talk. Douglas Manuel Yeah, agree. Jamie Brehaut Well, that's great to hear Doug with all his experience in dealing with the media and communicating to the public about complex healthcare issues. This is something that I've never been good at, and I really admire people who are good at that. Justin Presseau Yeah, I mean, what a unique talent Doug is that sort of stepping up in the public conversation during the pandemic with that and his concept of the community is my patient. That's just really, and maybe that's common parlance in public health physician space, but it just it I can see how that could be so motivating for someone in that position. And but also that plus the AI interface that, you know, there's not a lot of people that we have an opportunity to have both those kind of areas of expertise and how he navigates that is just yeah, really yeah, just a treat to hear him think through those things. Jamie Brehaut A couple of things that caught that came to mind or that sort of stuck in my mind were this notion of, which we all know, but it serves reminding, is that introductions of new technology almost always increase inequities. I think that's I think that's particularly likely and particularly something to consider when we're adopting these new tools is, well, let's think specifically about the potential for increasing inequities with these sorts of things. Yeah, I mean, the I hadn't thought of the AI scribe as being something if it's not your first language, then the scribe isn't going to scribe as well, and therefore the notes aren't going to be as good, and maybe there's more likelihood to to miss things along the way. Yeah, really seems a really important issue in the process of AI adoption. Justin Presseau It also made me, that that point, I tried to tie that back to how does that apply, you know, it's clearly applies to the healthcare space, but how does that apply to the general misinformation ecosystem and who is who are the early adopters and have access to the tools that could then be creating AI slop or creating misinformation through these that there's that kind of layer of inequity of access that positions others who don't have access to the tools and or even are interfacing with the space itself in an inequitable position. Jamie Brehaut Yeah. Yeah, and he also talked about, we were talking about trust and trust in physicians, but he talked about trusting the agents, the idea that that trust in the AI agents is also going to be an issue and maybe that's something we need to think more specifically about that is are people becoming too trusting too quickly or are they in some situations are they not being trusted enough? Yeah, I like I like that idea of considering the the extent to which trust in new AI tools is an issue that needs to be managed. Justin Presseau The other thing that I jotted down as well as he was speaking is that that concept it was go time during the pandemic for him to step up and having been on the receiving end of things and thinking about, you know, what would Doug think about this next this decision we have to make in the pandemic, that I feel like we we would so benefit from that voice now in a different context, clearly than outside of a pandemic emergency, but the we don't have that counterpoint right now in society I feel around the onslaught of misinformation and that is being turned out that those voices that at least some people like myself are gravitating to like I can understand the evidence, but hearing it distilled in a conversational way on the radio by Doug during the pandemic, it's just so helpful and that, you know, it's a lot to ask of an academic, a public health physician to continue in that kind of role in perpetuity, but the I don't know, it just struck me that there's something to be learned about that role, that trusted role, and that is seem it just feels a little missing right now. Not and it's not at all about Doug, but more about just that general voice that was prevalent in the pandemic is less so now, I feel. Jamie Brehaut Not just about Doug, but we need more Dougs. Justin Presseau Exactly, exactly. This podcast is produced in collaboration with PodCraft Productions. For links to the research discussed in this episode, go to the show notes. If you enjoyed this episode, we'd love your support. Subscribe wherever you get your podcasts, leave a review and a five-star rating. It helps others find our show. I'm Justin Presseau. Jamie Brehaut And I'm Jamie Brehaut. Thanks for listening.