Episode #356: The Rundown (Mattie Rogers' in the hospital, CRISPR, NBA and Achilles injuries, and More)

23:21Dr. Austin Baraki:But then you need it, you need it, you know, adequate insurance coverage. You actually need enough like cognitive, you need to be able to understand these devices and how to use them and how to troubleshoot them. So some patients are deemed like ineligible for some of these things on that basis. So there's still, it's still a tricky diagnosis to navigate for now. But if we can restore people's pancreases in the future, I mean, that'll be awesome. And I think we will. Yeah. I also wonder if there's going to be some application, maybe even to those with type two diabetes, because as you know, you know, after diagnosis and onset, it's almost like a ticking time bomb where the pancreas is kind of pickled, which is a medical term, obviously.

23:57And so I wonder, I mean, if it's a stem cell derived therapy, I'm thinking it's like, it says, Hey, hey, islet cells, why don't you get back to work, man? Wake up, bro. And they're like, Oh, shoot. Yeah. Dang. I knew I was, I thought I was late for something. Yeah. Yeah. So it'd be cool to see what other applications are for either this agent or similar agent or whatever, but kind of cracking this code as far as how to target those cells. That's a big deal. Yeah, I agree. All right. We're going to shift over to the NBA. Austin, you an NBA fan?

24:25Dr. Austin Baraki:From time to time, I wouldn't call myself a hardcore NBA fan, but I generally am interested in lots of different sports and respect the athleticism. Watching a bit of Wimbledon right now. The Tour de France just kicked off yesterday, so paying attention to that as well. Yeah, yeah, yeah. I'm more of a college basketball fan than NBA, but yeah, it's still amazing what these folks can do. Anyway, the Oklahoma City Thunder, well, they recently clinched the NBA championship in a Game 7 win over the Indiana Pacers. Now, in that game, Pacers star point guard Tyrese Halliburton tore his right Achilles tendon, making him the seventh player in this NBA season to go down with an Achilles injury.

25:02Now, seven in a year may not sound like a lot, but remember, the NBA only has about 450 total players. And additionally, between the years 1990 to 2023, a total of 45 NBA players have sustained an Achilles tendon tear or a rate of 1.36 per year. So seven is much, much higher. So the question to you, Dr. Baraki, is why are Achilles tears so common now in the NBA? I don't know.

25:32Dr. Austin Baraki:All right, moving on. Yeah, next question. Yeah, this is the injury, as we've talked about before, super multifactorial. I would actually be interested. I don't know if you want to ping, you know, Dr. Miles and see if he has a thought or a take on this based on his understanding of this condition, rehabbing folks with it. It's pathophysiology. But I think that in general, as with most, you know, acute injuries like this that we see, be it tendon tears, ligamentous injuries, muscle tears, things like that. A lot of it has to do with some sort of imbalance between the, you know, the training or the playing stress that's being put on the person and their kind of recovery ability outside of traumatic situations.

26:06Dr. Austin Baraki:And so, you know, I don't know enough to say is there more traumatic situations happening, for example, like you could imagine a scenario where there's a rule change allowing certain things that lead to more, you know, traumatic, you know, bumps between people that can increase this risk versus just the overall playing load, the amount of time on court, the amount of the amount of games per season, which as far as I'm aware, has not significantly changed but it makes me think for example about like how in major league baseball within the past few years there have been several rule changes that were deliberate to try to make the game move along a little bit faster and to make it a little bit more exciting so if i recall they like made the bases a little bit bigger to encourage more stealing and so there's some potential traumas that can happen when more bases are stolen and certain kind of slides are attempted and things like that and then they also instituted the pitch clock and so now you know i more recently i'll see a baseball game on TV, you know, if I'm someplace and I'm like, man, that pitcher is like immediately getting set up and ready to throw just because I'm so used to having grown up with baseball, where it was just like, super lackadaisical and slow, and they would take a ton of time between pitches.

27:08Dr. Austin Baraki:And now there's like almost no time at all, I can imagine that there is some sort of impact as far as that stress on pitchers, outside of just raw pitch count, and how that might impact their risk of aches, pains, injuries, things like that. So those would be the kind of areas where I I would look first to try to get a sense of an explanation for this, be it time on court or shots or traumatic things, things like that. But would be interested if you have any thoughts or if, again, if, you know, Derek or anybody else on our team has any thoughts on this would be interesting to hear. Yeah, I did ping Dr.

27:36Miles just to see if he had a take on this. So if he gets back to me before this podcast goes up, you'll hear that audio now. And if not, it'll just be two guys speculating about this. So a couple of things I looked into, like one, I was like, well, maybe like advancements in training, medicine and maybe financial incentive, whatever, has made the average age of the NBA higher. And so maybe this is like some age related contribution. Well, actually, the average age in the NBA is trending lower. So I kind of like, all right, well, that was fruitless. And then I thought about same thing as you talked about.

28:05Maybe the playing load is too high, which would include practice, competition, whatever. Now, the length of the season hasn't gotten any longer. Right. Right. And it's challenging to know how much playing time people are actually getting and further like what is the playing load? Is it minutes played? And does that accurately capture like how many shots they took, how many rebounds, whatever? You know, there's because just playing is different than like actually had more rebounds, less rebounds. You ran the court more or less like it's just challenging. Yeah.

28:34Dr. Austin Baraki:And it probably varies a ton by position. You know, what are the demands on each player per minute based on what position they play? That's going to be really variable. You have to you'd have to stratify it by that, too. Yeah. So the three things I came up with that are my leading three theories, and it's unlikely to be just one smoking gun. I think it's probably multifactorial. That's just a hedge and sort of doctor speak. Like, well, it could be multiple things. I think that there are a number of injuries that are happening and people are returning from those injuries too soon due to the incentives to do so, whether it's potential glory from like a postseason birth or something like that or winning the playoffs, something like that.

29:14But if people come back from two injuries, they're more vulnerable to re-injury, which would make sense, particularly if it's like an ankle injury, foot injury, other lower extremity injury, and perhaps that increases the risk of tendon rupture. Uh, the game has certainly changed over the last few decades. It's more open now with less defense. So there's more running. And so perhaps a similar total, like duration of time spent on court. Well, now maybe the actual load is higher. So I kind of favor those two things. And then also, and I actually didn't, you know, there's no real data on this, but I would like to see a lot more people are wearing low tops than, than high tops.

29:49And, and, you know, there's, I think there's a cadaver study looking at like the tensile force of the of uh that's imparted upon the achilles standing with low top versus high tops and yeah it's higher on low tops so perhaps some like mechanical type type contribution because low tops are maybe people like the way they look or they like the freedom i don't know me personally

30:08Dr. Austin Baraki:if i'm playing basketball i don't know that the footwear matters that's not your limiting factor yeah yeah i'd be curious if it's more a stylistic preference or if players feel that like maybe the mobility that they have in lower lower top sneakers is better but maybe that comes at a trade-off of a little bit more of that Achilles stress. Hard to say. Yeah. Also, can't rule out the possibility that it's just artifact. Like, eh, sometimes you're going to have a bad year. Yeah, maybe next year it'll revert to the mean. You know, that'd be interesting. Or lower. Yeah. There was no Achilles injury in 2020.

30:35Yeah, anyway. This podcast is brought to you by

30:37Dr. Jordan Feigenbaum:ButcherBox. I order groceries online constantly, which means I've made peace with a certain amount of disappointment. The produce always tests you, and the meat is a gamble. Half the time I open it and think, well, I guess this is dinner now. ButcherBox is the opposite of that. They deliver grass-fed beef, organic chicken, wild-caught seafood, and more, all sourced to standards that they're transparent about. Every box is customizable, so you pick the cuts and proteins that fit how you actually cook and what you like to eat, and you can change it up whenever you want. They've already done the sourcing homework, so the only thing left is deciding what you're making that night.

31:07Dr. Jordan Feigenbaum:And it tastes awesome. I've been grilling a lot, and the quality is legit every time. It shows up at my door, it's ready when I want it, and I know it's good. Whether you're grilling out on the weekend, you're feeding the family on a Tuesday, or just skipping another takeout order, you've already got something great ready to go. Go over to butcherbox.com slash barbell to get $20 off your first box, plus your choice of a free ribeye, New York strip, or filet mignon in every box for a year with free shipping always. That's butcherbox.com slash barbell. Be sure to use our link so they know we sent you.

31:36Dr. Jordan Feigenbaum:Have you ever noticed that you end up wearing the same few items on repeat, or is it just me? Now, I work from home most days, but I still get ready to head to the office every day, which fortunately is just down the hall. Anyway, last month, I finally noticed I was wearing the same items over and over again because, well, they just work well. They make getting dressed easy. That's what Quinn's does best. Their organic cotton mesh stitch sweater polo is one. I have it in almost every color. And the vintage wash camp shirts are another. That's been in heavy rotation all summer because I think it looks better than a t-shirt and it's just as comfortable.

32:08Dr. Jordan Feigenbaum:They also do premium pants, fitted tees, and polos, plus breathable activewear. Everything is priced 50 to 80 % less than comparable brands because they work directly with ethical factories and they skip the middlemen. And it goes past clothing. They do bath and kitchen essentials and even hotel quality bedding, which I also bought before they even sponsored this podcast. So find your next fall favorites at Quince. Download the Quince app for app exclusive offers or go to quince.com slash BBM. You can get free shipping on your order and 365 day returns. Now available in Canada and in the UK too.

32:40Dr. Jordan Feigenbaum:That's quince.com slash BBM. This podcast is brought to you by FIGS. Here's a piece of medical history that we didn't learn in school. Before the 1920s, pretty much everything in the operating room was white. The attire, the walls, everything. And the logic was cleanliness. The problem is that surgeons spent hours focusing on red tissue under the bright OR lights, and then their red photoreceptors fatigued, and every time they would glance away from the white fabric, they would see a blue-green ghost of whatever they were just looking at. So surgeons were getting headaches and seeing phantom anatomy mid-operation.

33:16Dr. Jordan Feigenbaum:But the fix turned out to be pretty simple. Surround the surgical field with the color that sits directly opposite red and it cancels out, which is why scrubs are green and blue today. Figs took the same approach with scrub design. They are innovative. They're lightweight, they're actually tailored, and they're antimicrobial. Plus, they're also built to survive the laundry. So no more boxy, big, scratchy scrubs. Figs rethought them completely with better tailoring and tons of styles and colors to choose from. And it's not just scrubs. They've teamed up with New Balance for supportive sneakers, they make compression socks for long days on your feet, and they have outerwear for when the hospital AC is on full blast.

33:51Dr. Jordan Feigenbaum:If you want to check them out, Figs is offering 15 % off your first purchase. Just go to wearfigs.com and use the code FIGSRX at checkout. That's wearfigs.com, code FIGSRX for 15 % off. 220 minus your age. It's on your watch, it's on the treadmill, and almost nobody can tell you where it came from. Well, I will. It came from a review paper published in the 1970s by Fox and colleagues looking at physical activity and heart disease. Here's what 220 minus age actually is. Picture a graph with age along the bottom, max heart rate on the side, and a cloud of dots from a bunch of different studies. 220 minus age is just a straight line that the researchers laid over that cloud.

34:29Dr. Jordan Feigenbaum:It wasn't calculated and it wasn't tested against anything. Someone just looked at the scatter plot and drew a line through the middle of it and said, yeah, that looks pretty good. That's the whole origin story. And the data in that original paper didn't even support it. The author said, and I'm quoting, no single line will adequately represent the data of the apparent decline of maximal heart rate with age. They said 220 minus age just defines a line not far off from many of the data points. That's the pedigree. And 50 years later, it's the default on every smartwatch sold. And it's not even the best formula that anyone's come up with.

35:00Dr. Jordan Feigenbaum:The better fit is 208 minus 0.7 times your age. It still carries about a 6-beat error, so what do you do? Like, if I'm going to program your conditioning, the easy runs, the cross-training, and the threshold work to make it faster for 16 weeks, I want to make sure the prescription fits you, not just a line that somebody drew through a cloud of dots. That's why I built a field test into the hybrid 5K and 10K templates, at the start and again at week 8, because fitness changes. Any heart rate estimate can be off by enough to throw off your intensity and your subsequent training load, which matters a lot when you're combining lifting and running in a way that makes you stronger and faster at the same time instead of just burning you out.

35:37Dr. Jordan Feigenbaum:So you get a 20-minute field test to make sure that you're anchored to the right intensity. The template includes three 16-week programs, 5K beginner for those who haven't done much running before, and the 5K and 10K performance versions for those who want to go fast. All of the programs include three days per week of lifting with every conditioning workout fully written out. And it's available now on our website, barbellmedicine.com, where you'll get both the desktop files and the app version of the program. The app is free, as it always is. Just search Barbell Medicine in the App Store. And like all of our other programs, it's included free for the Barbell Medicine premium subscribers.

36:10Dr. Jordan Feigenbaum:So if you want to get better at running without giving up your strength, check out our latest template, the hybrid 5K and 10K program. Okay, back to the show. All right, moving on. The American College of Cardiology is now recommending anti-obesity medications like semaglutide and terzepatide as first-line treatment to reduce the risk of heart disease. Now, in their expert consensus statement on medical weight management for optimization of cardiovascular health, that's a mouthful, they state that modern obesity medications have proven more effective than lifestyle interventions alone at not only weight loss, but at reducing overall cardiovascular disease risk, with fewer risks than procedure-based interventions like metabolic bariatric surgery, which we've covered on this podcast before.

36:54Now, prior guidelines have recommended trying lifestyle interventions alone before beginning medications. But this new guidance recommends considering these medications as a first-line treatment option for eligible patients. Dr. Olivia Gilbert, she's the chair of these clinical guidelines and a cardiologist at Wake Forest, well, she said that patients should not be required to, quote, try and fail lifestyle changes prior to initiating pharmacotherapy. Nonetheless, lifestyle intervention should always be offered in conjunction with obesity medications. Any thoughts on this update, Dr. Baraki?

37:28Dr. Austin Baraki:Yeah, honestly, something I've been wanting to do for a long time, maybe we can collaborate on this, is basically draw the highest quality data that we can on these anti-obesity drugs, which I think that that's even a bit of a narrow classification given how many health effects that they have, to just call them that. But pull together the highest quality data on these medications, potential benefits, potential risks and costs, et cetera, and then pull together the highest quality randomized trial evidence on lifestyle interventions in a real world context, not where you're like locking people up in a metabolic ward and force them to do things, but just like you institute, you know, the type of guidance that maybe fitness influencers really want doctors to do, the things that they say your doctor's not doing.

38:16Dr. Austin Baraki:How effective is that? Potential benefits. What are the potential risks of that? Costs, et cetera. And then kind of, I don't know if anonymize is the right term, but basically like blind it to what the actual intervention is. Just show the raw numbers and offer it to people. Which would you choose? Would you choose A or B without knowing what the interventions are? If you get this potential benefit, this potential downside, this cost, that potential benefit, that potential downside, that cost. which makes more sense to use or ideally use them together rather than forcing one or the other, right?

38:53Dr. Austin Baraki:I think that it would not be a terribly difficult decision for most people if they're just looking at these things. However, as soon as you actually are unblinded and you know what the interventions are that people are talking about, suddenly we get this really frustrating discourse to have to navigate out there where people do feel like everyone should try lifestyle first. And it's like, that's fine. I think that the position being recommended here is not that we force everyone onto these medicines, no matter what, but rather that they are a feasible option up front together with lifestyle interventions that you can offer people as part of what we'll call a shared decision -making process.

39:31Dr. Austin Baraki:rather than gatekeeping the medicines and saying, you cannot enjoy these potential benefits until you prove to me you earn the right to use these by not having success with Lifestom. It's a very perverted kind of way of approaching this. I think that obviously there would need to be discussions around the cost aspect because there are tons of people who are potentially eligible for these. and at current costs, you know, it is something that is unlikely to be sustainable for the healthcare system. Those types of calculations have been done to show like at what price point do these medicines become more worthwhile as far as like saving more by preventing these complications than they cost up front.

40:17Dr. Austin Baraki:And I actually don't think that we're quite there right yet as far as the costs in the U.S. go. These medicines can be had much more cheaply in other countries. And there it's like actually a no-brainer because of how much morbidity and preventable death that it prevents at relatively low cost. So those are my thoughts on this is I think it's totally, super, extremely reasonable to offer up front along with recommendations for lifestyle management instead of forcing people, as that physician said, forcing them to try and fail before you like earn the right to have some benefit. And this is especially in the context of what we know about cardiovascular risk as a cumulative lifelong kind of process.

40:56Dr. Austin Baraki:So if you force people to do the lifestyle and fail first, well, maybe that how long does that need to happen for? Is three months enough? Is six months enough? Does it have to be a year? Well, if they fail after a year, that's a year of exposure to obesity, to elevated blood pressure, to elevated blood lipids that could have been prevented and addressed a lot sooner. And there's a kind of a magnifying downstream risk at the end of their life that you are, you know, cutting out by treating them earlier. Yeah. Yeah. I mean, again, I think if you can, if you did like a, you know, side by side comparison, again, that was blinded, you know, for folks, they'd be like, wow, this option C, which in that case would be combined anti-BCA medications, which maybe need a rebrand and lifestyle stuff.

41:36You're like, wow, that's way more powerful than one, one or the other. It does get more interesting. I think when you have to pick between just lifestyle and just medications. It's, I don't know, like the best data we have on lifestyle stuff alone for weight loss, for example, it's just not, it's not that good. And as a person who started a kind of fitness and health company, like, I don't want it to be that way. I want to be able to tell you like, look, no matter who you are, where you come from, what your background is, what your resources are, we can help you, you know, a lot. And while I do think we can provide some benefits, exercise improves a lot of outcomes, for example, independent of weight, changing your dietary pattern to be something more health-promoting can improve outcomes regardless of what happens to your weight.

42:24All of that is true. But if I wanted to, quote, optimize somebody's health trajectory and reduce their risk to the maximum amount, it's probably going to require some pharmacotherapy. It may not necessarily be an anti-obesity medication, but it might be a blood pressure medication. It might be a lipid lowering type medication, you know, whatever. We can use both as, as you have, and I've kind of co-opted this, this quote, you don't get a gold star for going through life without any help. Yeah. You know? Yeah. So I thought this was, you know, an evolution of kind of the, the grumblings and the thoughts around these agents and where they should be used.

43:00I really didn't love the idea of try lifestyle first and fail it. And then you, we can add some help. It's like, why would you do that? You don't do that for blood pressure medication, right? Especially if somebody's like, it's been elevated for like a while, because you want to get it down to normal. So people reduce that exposure immediately. Same thing with lipid lowering, same thing with blood sugars. I mean, whatever, what other condition are we like? Yeah, let's like, let this thing persist potentially for an unknown amount of time and make sure that you fail first. And I want you to feel bad about it.

43:29And then we'll, we'll give you a helping hand.

43:31Dr. Austin Baraki:You know, here's an interesting, more hot take than tepid take is if the fitness folks, the influencers here who, you know, cry and moan about this all the time, if they wanted to be logically consistent, then they should also be telling people, you should try to do all of this health, fitness, diet, exercise stuff on your own first. Yeah. Fail at all. Before I am willing to help you because I'm like a higher level of intervention or a more intensive intervention, just like the medications are a more intensive intervention. And so I will not actually coach you until you prove to me that you need coaching by failing on your own.

44:06Dr. Austin Baraki:That would be logical consistency. I don't think that we're likely to see that happening. Somebody slides into their DMs and they're like, hey, I'd really like you to coach me. He's like, oh, well, have you tried it on your own? And someone's like, no. And you're like, no. So I need you to do it for a year on your own and make sure that you fail as evidenced by these criteria. And then I'll consider helping you. Yeah, yeah. That's the logical consistency. So let's see it. Yeah. All right. Hey, speaking of heart disease, did you know, Dr. Baraki, that heart attacks are no longer the leading cause of death in the United States?

44:37I did since I sent this to you. That's true. Well, based on data collected from 1970 to 2022 using the U.S. Centers for Disease Control and Prevention's WONDER database, this tracks all recorded fatalities in the country. In 2022, heart attacks accounted for 24 % of all deaths in the U.S., down from 41 % in 1970. This decline is largely thanks to an almost 90 % decrease in heart attack deaths, which were once the deadliest form of heart disease. Is this an incredible breakthrough or are we missing something?

45:11Dr. Austin Baraki:No, this is an incredible breakthrough, but also I don't even actually call it a breakthrough. This is also unsurprising to me. Um, this is something that, uh, I have expected and was predictable for, for many years at this point. I think that heart attack or acute myocardial infarction, where you have a sudden, you know, atherosclerotic plaque rupture and you get a clot in one of your coronaries and insufficient blood flow to your heart muscle and the heart muscle dies unless you get a heart catheterization and potentially a stent put in as soon as possible. Um, that is kind of the situation that we're talking about.

45:45Dr. Austin Baraki:And the incidence of that has indeed been going down, in particular, deaths from heart attacks has been going down as well. And this is because we're getting much, much, much better at a couple things. We're getting better at prevention. And so that relates to all of the risk factors that we talk about all the time, especially blood lipid, blood pressure, and most, and very importantly here, smoking-related issues. And then additionally, the actual treatments for this when they do happen, when these heart attacks do happen and patients show up to the hospital, we are far better at treating these things now than we were in prior decades.

46:19Dr. Austin Baraki:This is both due to advancements in the technology of intervention and the stents and interventional cardiologist skills, and also the medications that are used around and after that procedure to prevent the risk of death and complications from it. And so that kind of very acute event and death from it, that death has declined from that, very unsurprising to me, but the other concern is what replaces it as far as the leading, you know, causes of death. And instead of the much more acute, sudden kind of heart attack induced death, now it's much more chronic heart disease related issues, especially things like heart failure.

46:56Dr. Austin Baraki:And for me, you know, if you took someone like me, an internist who had been in, who's, you know, actively practicing in the hospital setting, you know, many decades ago, they would be seeing heart attacks left and right all the time because of how common they were. I actually, you know, I still see them from time to time, but they are not like an absolute daily occurrence that I'm seeing all the time. Whereas the more chronic forms of heart issues like heart failure, I do see near daily all the time. So we're preventing the much more acute heart attacks and the death from them, but rather patients are left dealing with the consequences of it longer term with like weakened hearts, for example, as a simplified way to think about it for very long periods of time.

47:39Dr. Austin Baraki:Now, even the treatments and the management of chronic heart failure has improved a ton such that, you know, death has improved. That decline in the rate of death from heart failure has actually unfortunately plateaued a little bit in recent years and hospitalization from it. And I think a lot of that has to do with kind of uptake of some of our best medicines that we have to manage it, as well as like the increasing burden. Patients are sicker, more complicated. They're living longer with more medical conditions. They're on more medicines to manage these things. And sometimes it can be very challenging to handle these situations.

48:10Dr. Austin Baraki:But the short story is, not surprising, the very acute death decreasing, the chronic heart disease continues to rise, and it's a pretty significant burden. So that's what I spend a ton of my days actually managing is the more chronic forms. Yeah. And the American College of Cardiology actually just recently updated their heart failure sort of scientific consensus and a similar sort of thing to what they were talking about with heart disease, generally speaking. They're like, look, instead of just lifestyle alone, trying that and failing it, maybe let's use some of these anti-obesity medications because they recognize the potential benefit there as well.

48:45I do suspect that we'll get to some point, we'll just keep kicking the can down the road. I do suspect that some type of chronic heart disease-related thing is still going to be the leading cause. Something's got to be the leader. Yes, always. And I think it's going to be some sort of chronic heart disease, but at some point it's not going to be heart failure. At some point it's not going to be like hypertensive related, you know, cardiovascular disease or related to lipids. We're just going to catch all that stuff and have great interventions and there's going to be good uptake. It's going to, you know, just kick the can down further and further and further.

49:17I think if at some point it's like, yeah, these are just structural abnormalities that we couldn't pick up. Like that would be a great place to end up getting, you know, because it's like anything that we could reasonably like prevent, predict, identify and intervene upon. We've nailed that. And so now we're in a more even more complex sort of situation. Like do you screen newborns for these like, you know, structural abnormalities? And if so, like how do you how do you fix them? Yeah. So we'll see if we keep we're doing this another 10 years from now. Maybe we'll have an update. But I don't know.

49:49All right. Last topic here on the second edition of the rundown. More AI news, since you're our resident AI and medicine expert. Well, you know more than I do, in which case that makes you the expert. Yeah, like I said, third best. Third best, that's right. I don't speak Italian at all. So Microsoft is claiming that their MAI diagnostic orchestrator, this queries several leading AI models, it's saying that they can diagnose patients more accurately than human doctors. This is a preprint, hasn't made its way into peer-reviewed research yet, but I suspect it will. The Microsoft team used 304 case studies sourced from the New England Journal of Medicine, And in their experiment, either a physician or their AI begins with the short case abstract, and then they iteratively request additional details from a gatekeeper model.

50:44Basically, they ask a discrete question, get an answer. They may run a test or ask an additional question. They keep going back and forth. They're saying their AI model outperformed human doctors, achieving an accuracy of 80 % compared to the doctor's 20%. It also reduced costs by 20 % by selecting less expensive tests and procedures. So the question is, are we cooked? I'm going to say not yet, with a heavy emphasis on the yet. So I looked a little bit at this news story because it obviously intrigued me.

51:20Dr. Austin Baraki:It is something that I do day to day is diagnose, right? That's what we do with these mystery cases. And if an AI can just very easily handle these things, then yeah, it might be cooked, man, as to your point. But there were a few issues that I saw with this particular study, as well as some of the conversation around it among physicians. Those case studies in the New England Journal of Medicine, I read them every week. I think they come out every Thursday. They're excellent mystery cases. There are a few things about them. They are, in general, challenging. They're not putting the simplest things there.

51:59Dr. Austin Baraki:Pretty specialized too, also, just as an aside. Yeah. I mean, that's a big part of where I nerd out on these things and how I have gotten better at doing this is by reading and learning from those things every week. The other things about those cases are that they are published. And there is, I don't know if it was reported, whether these AI models were trained on that material. So if they were, then they already knew the answers effectively because they were trained on that material, right? So it wasn't truly a mystery in that sense. But that aside, all those cases are published because they have a clear diagnosis and conclusion.

52:35Dr. Austin Baraki:And for anyone who is a regular listener, including to our last episode where we said sometimes medicine is messy and there's not always a clear answer, or even the TIA conversation that we had earlier today, where things don't always have a clear, concrete diagnosis at the end of the case. And so that does impact things quite a bit if all they were using were cases that had a concrete diagnosis. For example, there were no just like completely healthy patients. All these people had a final definitive diagnosis. Sometimes people have a headache and it's just like, you know, a stress-related headache.

53:07Dr. Austin Baraki:It's not an underlying, you know, brain tumor or something like that, right? So there were no essentially like healthy cases mixed in with this that it had to try to differentiate, right? It was only on these kind of rare diagnostic zebras that we talk about, where you keep testing and testing and testing because there is a final diagnosis. At some point in clinical practice, when you're seeing patients, at some point, a physician has to make an assessment of, at what point is it the correct answer to stop testing for now, maybe, to stop searching, and to give things a tincture of time, as we call it.

53:43Dr. Austin Baraki:And a lot of things do tend to get better on their own with time. So instead of all these already solved, published cases on which it could have been trained with no uncertainty, with no healthy patients in the mix, that differs a ton from actual real-life practice. The other thing is that the physicians in the study who were actually trying to solve these cases, they were not allowed to use the internet to search. They were not allowed to consult colleagues or consult specialists. They were not able to use up-to-date or medical reference resources, things like that, which is also completely unrealistic.

54:23Dr. Austin Baraki:Like, just because I've gone through school and training and have a fair amount of, you know, postgraduate experience at this point in my own practice, I still am using those kind of medical reference resources most days in practice, because that is the reality of what needs to happen, right? So I can't have the entirety of medical knowledge in my brain. Now, that might be feasible for an AI, but at the same time, there are other things that I have that an AI doesn't and vice versa, right? Right. And so the other aspect is, you know, you mentioned this difference in like a kind of cost effectiveness, and it was just using like one relatively limited measure of cost effectiveness, like just like the dollar cost of the test, for example.

55:05Dr. Austin Baraki:But there's a lot of other costs and downsides to recommending more and more and more and more testing to patients that we have talked about before. And when you know that there is an answer at the end of the road, you will probably keep doing more and more and more testing to get there. Whereas again, in real life, where there may not always be an answer at some point, you have to say, well, it's actually not, there's not ongoing benefit to keep doing more testing. Maybe we've ruled out all the bad things. Maybe we give this a little bit of time, some amount of time, some days, some weeks, some months, something like that, and see if it gets better on its own.

55:35Dr. Austin Baraki:And again, a lot of things do. I remember I often quote this line in one of my lectures on this topic. It was from an Annals of Internal Medicine paper, where it said something to the effect of the vast majority of symptoms defy definitive biomedical diagnosis. And a lot of them have a favorable natural history. That is not the type of case that was being done here. So I think that, you know, at the risk of sounding a little bit like a complainer, I think that this test was rigged against the physicians. And yes, exactly. Hate using that word here, but and then in favor of the AI. In other words, it was set up for the AI to have as strong of a chance as possible and the physicians to have as weak of a chance as possible, right?

56:13Dr. Austin Baraki:physicians isolated, only what's in their brain, no resources, no consultation, no specialists, no conversation, so not reflective of real-world practice. And for the AI, cases that had definitive proven diagnoses, no healthy patients, no counseling involved to say, let's take a pause and come back and follow up and things like that. So I mean, I think that this technology has obviously improved really dramatically. It's not surprising to me that by collating all of this expert knowledge that's contained in those New England Journal cases, among other places that it can perform phenomenally well here but there are additional layers to this to be more reflective of real world practice so it's going to need to get trained and tested on like real patient populations that include a lot of healthy people right or people who do not have a definitive underlying you know embryonal rhabdomyosarcoma or something like that but people instead have like uh some non-specific abdominal pain that gets better and doesn't actually represent anything serious, right?

57:09Dr. Austin Baraki:And then I think if you're going to be assessing cost-effectiveness, you also need to be assessing harms and risks of over-diagnosis and making your comparison more realistic with actual physicians who are in real-world practice who have the access to resources that they are already using and specialist consultation and things like that. That is how the real system works. So those are my thoughts. I think that it's, you know, definitely something that's going to continue to progress and may well end up overtaking a lot of us in some way, particularly in non-procedural specialties. But those are some of the issues that jumped out at me and I saw, you know, discussed around this as it related to, you know, whether we are in fact cooked or not.

57:50Yeah. I mean, I suspect that, you know, the combination of a trained clinician and AI together, like whether it's this model, this or another would be superior to, you know, that has been tested.

58:03Dr. Austin Baraki:That has been tested and it did not work out that way. Correct. Yes. And so again, I do think though, like how you would, you know, implement this in clinical practice and ultimately how you're testing the outcomes ultimately determines like how that, you know, what side, what side the results fall on. But yeah, if, if and when, probably more when than if AI is further refined to a point where we can reliably use this for either whether it's picking tests, diagnostics, workup, etc. It's just going to move the responsibilities of the clinician to a different point, which is fine. I'm okay with that.

58:39I don't know where that point is and like what the responsibilities of the clinician end up being, but it'll be different. And I'm excited to see what it is.

58:47Dr. Austin Baraki:Yeah, yeah, for sure. I think that there's still obviously going to be some human element in clinical medicine. I think people want that in general. I think that, you know, the worst case I'm imagining is from, I don't know if you remember seeing the movie Idiocracy when he sees the doctor and the patient just ends up going through this like robot machine with a bunch of probes that he has stuck in him. And then it just like spits out a diagnosis. And, you know, obviously it's a satirical kind of take on it. But, you know, the aspects of my job that I imagine might be a little bit more difficult to replace until we get some hyper-realistic humanoid robots that people deem indistinguishable.

59:21Dr. Austin Baraki:So like we're living in Westworld that can do a hospice meeting, an end of life conversation, guiding people through the dying process. Like those are things that I do day to day that are not just like rote memory of, oh, I recognize this set of symptoms and it suggests this diagnosis. Yes, diagnosis confirmed. Here's your treatment and you're good to go. There's a lot more to clinical practice than that. So, you know, one day, but hopefully I'll be on the outs from my career by the time that's the case. Retired, yeah. All right. Last thing, a little update here. uh do you watch any of the u.s open for golf uh very little but uh you kept me up to date on on some of the happenings but yeah what's up where our predictions were wildly bad yes yeah so uh i had picked bryson dechambeau to win just because i'm like look he hits the ball very far it tends to be a good putter particularly on weeks he's playing well uh this course sets up nicely for him he's on fire well he didn't even make the cut yeah did not even make the weekend so that that money that I had placed on him down the drain.

1:00:22You pick Scotty Shuffler. He finished a distant seventh. I believe he was like six shots back or five shots back. J.J. Spahn, first time major champion, end up winning this in dramatic fashion. If you're not a golf fan, this would have been a tournament to watch because it had all the drama. And yeah, I think as far as spectator sports go, golf might be one of the worst. yeah like uh and it just reminded me when you said oh look at the pitch there's a pitch clock in baseball and they're theoretically supposed to be like uh uh you know you have this amount of time to play your shot in golf good god they do not enforce this hardly ever and it is painstakingly slow to watch golf especially at the end of a tournament now i don't know if that's because there's only like a handful of groups left on the course.

1:01:15You know, there's like three holes left. And so you got what, three or four groups left. And so you're seeing like a handful of shots an hour and you're like, can we just, can we just go? But anyway, yeah, it was really, really interesting tournament. There's one more major left. It's the British Open, actually called The Open. And we'll cover that on the next episode of the rundown. But let us know what you guys think. These are kind of like topical conversations. I thought this would be an interesting addition to our normal podcast episodes where we talk about one specific thing. We do mystery cases.

1:01:49We do Q &As, all sorts of stuff. But yeah, if you guys like this, let us know. Leave a comment wherever you can, and we'll take that into consideration. But before you guys go anywhere, please leave us a five-star rating and a review. It really helps drive traffic to our podcast so we can keep bringing you all the latest nuance in health and fitness. From everyone here at Barbell Medicine, I'm Dr. Jordan Feigenbaum. We'll catch you next week and every week right here on the Barbell Medicine Podcast.

1:02:23We'll see you next time.