rounds. I'm Bob Walker, chair of the department and we have a special session today as people as people sign on. This will be a fireside chat with Dr. Topo. I'll be brief on Eric's bio if I wasn't it would take the entire hour. So Eric is the founder and director of the Scripps Research translational Institute, a world renowned cardiologist and geneticist. He is focused on using genomic and digital data to make health care better and an individualized and democratize health care. He's published over 1300 peer review articles as a member of National Academy, and is one of the 10 most cited researchers in medicine. He's authored three terrific best selling books on the future of medicine. He founded the new medical school at the Cleveland Clinic. He's the editor in chief of Medscape publishes the absolutely essential substack newsletter called Ground truths, and has been a go to person on social media since COVID, with about 700,000 followers and has done an immense public service by basically providing an annotated, very thoughtfully annotated, up to the minute literature review on on everything and COVID. Probably most importantly, he was a resident in medicine at UCSF, so many of the leaders I note in COVID word including Ashish Jha and Rob Califf, and others. And I have to say, Eric, when when, and if you can come on, that'd be great. When I tell folks who are of that vintage that, that I'm talking to you, they always tell me Eric Topol stories, and this is now 45 years later. So let's, why don't we start we'll get we're going to spend most of our time in COVID, a little bit of time on digital transformation and AI. But tell tell us what you were like as a resident love to hear this?
Well, Bob, it's great to be back with you at UCSF Grand Rounds. And thanks for revitalizing Grand Rounds, they're really been terrific. I've watched a number of them and glad to participate again. When I was there, from 79, to 82, it was a pretty, it was a fantastic experience. And I I thoroughly enjoyed I think it really set off my theory on good footing. But, you know, I got to see everything between the three centers. And, you know, the the the ability to have just not just from faculty and attendings, but from your your residents, to work with them is just a phenomenal experience.
And you've had really one of the most extraordinary careers, I can think of an academic medicine, when you when young folks come to you for advice, like, is there anything, any pearls that you've you've had over the years for in terms of telling, let's say, a resident or a medical student about how to think about their career?
Well, as far as the training part, you know, being at a place that you see it all, or as close as you can get to that, and have the people there that are really supportive? Make it fun? You know, that's, I think, really idea, I think there are a lot of training programs that are not as as much fun and not as much exposure to seeing as much as you possibly can. Medicine. So that's kind of the general guidance that I would provide. Great.
Let's, let's start with COVID. So maybe we'll start with very open ended. Tell us how you describe the current moment.
Well, it's about as good as it's been for a long time. You know, we got through this x BB one five, this kind of unique recombinant with two significant mutations on top of that, we got through the in this country pretty darn well. And we're standing that was a big deal because it was predicted from its functionality to be quite transmissible and potentially quite a bit of immune escape. So it looks as all our immunity wall that is both the vaccinations, the boosters, the infections and all their combinations is holding up well. Now, this is 15 months or so 16 months into Omicron. And that is a good thing because we haven't seen anything past Omicron. So unless we see something else, you know, we might be in a true endemic state. It's much better to diagnose that when you go backwards and say, Oh, wow, it's been relatively quiet with little wavelets nothing major sir no major surgery. There's really only two variants that are kind of taken over the world and it's another cousin of X bb 1519. Team, so, you know, things look good. The real question, Bob is, are we going to see that family unexpected, like AMR crime that is a pie or sigma, whatever you want to call it, where the virus finds yet another major new path to get even more transmissible, more immune invasive, potentially even more virulent. That's what is out there dangling all of us, of course, hope we never see that. But there's no guarantee.
And, again, it's hard, no guarantee and hard to know where things go. But is your sense that COVID as people think about behaviors, as as policymakers think about policies, is your sense that we're now living in the world COVID wise, that we're gonna be living in two years from now,
that might be the case. If we don't have another family of variants. This is kind of what we're stuck with for the foreseeable future. And that's why I still think we really need more tools to work with, like nasal vaccine, so you feel more confident about not getting infected. Paying Coronavirus vaccines and just in case that bad news family of new SARS cov. Two comes upon us. But yeah, and of course, what happens at PAX COVID starts to show any signs of resistance, which fortunately it hasn't. We're at a monoclonal antibodies, none of those work. We've left our immunocompromised in a very compromised position. So, especially for the people of advanced age, you know, it's it's not an ideal theme, because the virus will continue to circulate. And, you know, if you let your guard down completely, you could wind up getting COVID. And, you know, and then it's that unpredictable aspect of long COVID that we can't ignore. Yeah.
I'm in a hotel room. So actually in San Diego at a meeting. You have you've talked about this need for more and better research on vaccines and nasal vaccines and pan corona virus. Where do you think we are and on the pathway to them both in terms of the scientific enterprise, but also the policy and funding landscape
are really in bad shape for funding and priority? I'm hoping that we're going to hear more about that in the weeks ahead. That somehow the the US found some money to put towards it, but we haven't heard it yet. Now with respect to the science, it's exciting. It's just as exciting as the the original mRNA and other vaccines that came upon us so quickly. The idea is that we've already seen one nasal vaccine in India by Barrett biotech, which is a US Washington University, Michael diamond intellectual property. And we have many others Florian Krammer, and Akiko Saki. So many people in the US cite leading scientists have come up with very attractive nasal vaccine candidates, but they're not being pursued in this country. Or if they are, they're in slomo. And the silly thing here is that we keep thinking this is like flu. And flu mist doesn't work that well. But we already know that we can triumph over this virus so much better. I mean, look at PAX COVID. Versus Tamiflu are looking at our vaccines compared to flu vaccine. So I'm very confident that needle vaccines will work. We already seen one actually a new data from a second one, it looks quite good. And there are many more in the hopper. The point is that you're getting to Bob, we haven't given it the priority resources to succeed quickly. And, you know, it's a shame if we have to import these from other countries when it was actually our inventions. And I think nasal sprays are very alluring, because if even had to take a spray every three or four months, it doesn't have the systemic reactive genic effects. And it could really pronounced reduction of infections, which is what I think we we just don't want to promote and we've never since Omicron came along, we've never had a good way to and a durable way to have a marked reduction of infection propensity.
And with the science say that it is possible to have a nasal vaccine that would I've always wondered, Is it just you give it a spritz and it lasts a week or two? And so then what's the use case when to use it? But do you think that the science would say that we might be able to find one that lasts for three months?
Yeah, I think three or four months is probably likely it's possibly longer. But once you have it, then you have got a platform to try to get better durability. And that's the other thing regarding the shots, you know, who wants to get a shot every year or six every six months, which is probably more than what the data data tells us. You know, I know when I get I've had five and we're not every time I Get The Shot just so darn sick for a couple of days, I don't really want to go through that on a frequent basis. So why can't we get a more durable, variant proof vaccine by tweaking the the nanoparticles the mRNA, and making it effective against all beta coronaviruses, which is in our reach, we have so many monoclonal antibodies from the treasure chests of people, those rare people that make these potent rod neutralizing antibodies, which are the template for just these types of vaccines, but we haven't taken them across the goal line in terms of doing clinical trials, because we don't have the public private partnership, like we have with Operation warp speed. And the companies don't want to take the risk of of making it and doing clinical trials without support and de risking
it. Do you have any reason for it feels like in Washington for reasons that are quite understandable. They've kind of moved on in terms of investment in COVID, do you have any reason for optimism that they haven't and that they're going to invest in, in these kinds of techniques?
Well, there's certainly enough people in the science community that are clamoring that, you know, we're on the brink of success in these areas. Ashish Jha, who we both know very well, is still a major proponent who's pushing hard he's, he's, you know, been up against the wall. But I think it's pretty likely that there'll be some support, you know, but until we see it, and how, how much of it how much of a priority, my understanding is that there is real interest in moving forward. It's, it's, we could have had these by now, we create data fetch, but it's never too late. And the other point is, prepare for the worst case scenario. Because if it's something beyond Omicron, whatever that family name, Greek letter is, it's not going to be good. That means the virus has found yet another path to hurt us, and to find hosts and repeat hosts at at scale. So we need to prepare for always the worst case scenario. And right now, we've got more capitulation than preparation.
You mentioned that you, you know, the current vaccine by Val and does have some efficacy after six months, but it's clearly his wane a lot in terms of infection and a modest amount or moderate amount in terms of severe infection. Do you think it's waned enough that as we hit the six month mark, those of us who got the buy balance in September we should get another one or only for much older people? Or how do you think about that are versus a yearly?
Yeah, that's a really important question. A lot of people are asking I'm sure to you as well. The problem is that we treat everybody the same. And really, if you've had a prior COVID infection, so you have hybrid immunity with the vaccines, your durability, protection is much longer. But if you're like me, I've never had COVID nor nor I, yeah, no plan to get it. But I don't have that durability. So you know, that really good Lancet paper from the Washington University Washington group last week, I think helped put it put in perspective that people who get through COVID, without problems do have that feature of more durable protection. So you know, that idea of the by valen now is started in September, we're getting to the six month or more March 7 month, March, Mark. And this is troubling for people who never had COVID who are advanced age and there's no plan until next fall, do something about it. So, you know, I think it's a problem. There is some there is some vulnerability there that hasn't really been addressed.
So since you're only a few years older than I am, I will not call you of advanced age. But if someone's and you've already said get any vaccines you tend to have be a little have some reactions. If someone said Walgreens calls you this afternoon says we got your buy valent it's ready for you. We're gonna go in and take it are you gonna wait another six months?
I'll wait it out, because I think the virus is low circulating level. And if I saw a change in that, or if I was going to a place traveling, that there was a problem area, then I would probably sign up but I really don't want to get it when things are relatively quiescent.
So you the way you think about it is your risk has your risk in your vulnerability has gone up in this as you go hit six, seven months. But your chances of exposure seem low enough now that as you weigh those two things, you'd rather wait until the chance of exposure because the prevalence is higher. Change is not so much that you think your booster is still working super well, at six or seven months. Right?
I think the issue is, you know, protection against hospitalizations and deaths remained pretty high, you do get another small jump from the booster to keep it up at that very high 85 90% level. But we don't have any data for that six shot. I mean, in the past, we've had data from Israel and other places here, we don't have any. So, you know, it's kind of odd that you'd think oh, well, after the fifth one, the sixth one works differently. But, you know, it would be nice if we had some actual evidence, and we don't right now.
So at this point, you would wait, what if what if there was a pretty big surge in San Diego, you would get one now? Yeah, I would take the take the hit yes. Shot. You've done an extraordinary job annotating the literature, thank you. Because we're all dependent on that. Let me just sort of run through a few areas. And I'd love to hear your take on you know, I think people are following this less carefully than they used to, you know, for a lot of reasons, they may not be on Twitter all the time, the way they used to. So let me just give you a few topic areas. And if you could just sort of talk about what you think the literature has taught us in the last couple of months, sort of new studies that have changed the way you think about these topics. So we already started on one, which is the efficacy of the by valent. And how long it lasts. And, and, and how effective is both against COVID, against severe COVID. And also against Lon COVID. Anything else you'd add in terms of what we've learned about the this experiment? I think you made the point that when the by violent rolled out, we didn't have a lot of date on it. So big experiment, what have we learned in the last few months about it?
Well, I think by valen, performed better than was expected. Remember, as you well know about when it was released in September, we had no human data with that particular shot, a BA five booster. But when now we have lots of data, and it has everything about it, whether it's the lab for neutralization, whether in fact that we see cross reactive antibodies to x, BB one, five, and prior to that to BQ. One, one. So it's it's done a lot to broaden our immunity. And also the clinical data looks really good. So whether you had to have VA five, or whether just another booster. I mean, we can't say that definitively. But it does appear that having that VA five spike, which is half of these vaccine boosters that we've had has broaden the immunity pretty substantially, I don't think we would have seen that. We've just the Wuhan ancestral spike,
if you think we stuck with the original booster, there would have been more of a US about surge in cases with the new variants.
Exactly. Because you know, it's so different if you look at the antigenic distance between, you know, Wuhan Omicron and compare that from the original Omicron VA one to SBB, one five, I mean, this is much more distance. So I don't think that just having the original booster would have gotten us this far. And I do think it's really helped the protection. And we've seen it how well we did against the HPV one five wave in the United States for only the Northeast had a bit of a surge. But even that isn't entirely clear that it was related to this tough barrier. So overall, better than expected and give credit here because I questioned it, as you may recall. And I think, why don't we wait till we have more data, at least in the lab, but all the lab data once it once it was thoroughly reviewed, look quite encouraging.
And you're attributing some of the lack of a surge with the SBB and the new variants to the Bible. And but not that many people got the Bible. And so maybe take us through you mentioned the lancet paper, it's sort of the new literature on how well infection immunity works.
Right. So first thing about that is, you know, when you want to gauge about hospitalizations and deaths, you look to people 65 and older, and we did get 40% to get the buy valen it's not great. Remember, as you well know, we had 95% of Americans 65 and older got the original two shots. So 40% in great, but remember also that only 15% of the country eligible got the by valence 16%. So at least we had some better participation among people who needed it the most. Yeah, so I think that be prior COVID story is one that has been neglected by CDC. And it's really unfortunate had we acknowledged and I tried to plead with the CDC and other means that prior COVID was worth at least one shot. If we had done that, we would have decreased the polarization. Because ultimately, that's what the data showed, in fact, maybe even a case of prior COVID, particularly within the last year could be seen as protective as to shots. But we never gave it as in many other countries throughout Europe and Asia, where if you had documented prior COVID, that, you know, don't don't have to get the full series. And natural as it's called, or better infection induced immunity. Is, is important, because it's to the whole virus, not just to the spike protein. So it's really disappointing that we didn't we made these mandates, you know, you have to get the shots. And, you know, we just were aggravated already, you know, the, the anti Vax situation by not acknowledging that this was another way to achieve immunity and that, you know, together now, obviously, if you had never gotten a shot, we know that from various sources now that getting at least one vaccine, with prior COVID helps to prevent against long COVID. This is this, there are several recent studies to suggest that so there really is a lot more to hybrid immunity than we had originally thought.
Do you think some of the that was just the political environment was such that if you know that the folks that were sort of hoping to push vaccines for perfectly reasonable, you know, it was the right strategy felt like if they gave too much credit to infection, they were basically be saying, you're okay to get infected rather than get a vaccine. It's a tricky political dynamic.
It was tricky. But when I talked to senior management at CDC, I said, Why don't you do this? It's not that hard. They said, No, it's very hard, because we don't know if they really had prior COVID. I said, we'll only make it if they have a BS PCR test that you can review. And they just said, it's too complicated. So it was more about the convenience or lack of the inconvenience, I think, but if you go back, I think that mandate thing really has led to, you know, more of the war, or exacerbated the battles that, you know, we just didn't need to raise the heat like that, because there was something to it. And then the more data you look, the more clearer it became.
It's much influences your thinking now that if, if, like, if someone was eligible for the shot, if they'd had an infection, let's say in the last six to 12 months, you feel like that's the moral equivalent of the shot, they really don't need another shot.
Yeah, I mean, I think that they had they have they had, bonafide COVID. And they, they've got a booster at this point, right. And particularly since they got exposed to, you know, as close to the current circulating virus strain, as you can imagine. But the other thing you alluded to, which I think is really important is we have built in a very sloppy way, a very a relatively strong immunity wall. That isn't we didn't get people vaccinated and boosted as we'd hoped. But overall, the fact that we could stand up so well, to these recent variants, better than expected, is a sign that our immunity wall is, is relatively strong. wasn't the best route to get there. Right. But I think we can say in retrospect, it will be hard for Omicron family variants to unless, you know, there's some peculiar recombinant and, you know, very unusual mutations that occur outside this fight. It'd be hard for that family to further hurt us based on our current immunity wall.
Maybe the most controversial statements come in the last month or so is the Cochrane review of masks wherever you want masks.
Yeah, this is just crazy stuff to me. Because I just trying to understand how if you wear a mask, particularly one that has really good filtration, how it can help to inhibit a respiratory virus infection. To me, it's like a parachute. I mean, do you have to question that you need a parachute to jump off a plane? I just don't get this. Yeah, and I don't need you know, something. You don't need randomized trials before. And these are the kinds of trials that are virtually impossible to do. So when you have an analysis that tries to hold together evidence that comes out inconclusive based on the kind of evidence that that we might apply for devices and drugs and other things, but maths is just not in that category to me. And there's been, I think, some really good recent. You know, Trisha Greenlee from the UK, and others have taken apart the concrete and also put it in perspective, why maths, the data is so strong. So to me, it's just this is a no brainer. I don't know why this is continued to be question.
I said into the politics to get it fed. Don't make me do something.
Yeah, the mandate, if we just keep stop the darn mandates, you know, if people just you know, if you're intelligent, you realize that a high quality mask is going to help to some degree. Now, you could say, you know, what is the magnitude of that I'm accepting that it's not 100%. Even if you have a tight fit, you know, it's not 100%. But, I mean, why do we wear masks? In the hospital? Bob? Yeah, you know, helped me. You know, this is just crazy stuff
around there. You know, there are a lot of them are currently studies, and a lot of them are studies of mask mandates, which is a different question. To our people. Do people wear masks? And do they wear good masks? And they wear them correctly? It's a different question. But people saw that and say, Oh, I told you masks don't work, which of course is crazy land.
Yeah. Interesting. You have one of your colleagues, their faculty members of the UCSF who are leading the charge against all the things that we've been talking about as well,
we won't go there. What have we learned about Lancome in the last couple of months?
Yeah. Not as much as I had hoped we would. So this is a serious issue. That is the enduring part of the pandemic that is the inconvenient truth, where we're long on mechanisms. We're short on treatments, ie none, we have no definitive treatment for any component of, of long COVID, of which there are multiple, you know, on the one hand is dysautonomia with postural orthostatic tachycardia. Now there, there's the more pure immune aspects, whether that's manifests through brain fog and many other symptoms. And we haven't made headway. I think we've made some good there's been some good science to understand and and basically, backup the mechanisms of a heterogeneous symptom complex. But we're, we've left these people in the lurch, we haven't come up with anything to help treat them. And it's a serious matter, because we have around the world 10s of millions of people who are suffering. And what's really bad is in the medical community that not taking this seriously, or in terms of dismissing it in patients. And also the idea that we haven't gone after it in terms of the randomized trials here. We do need randomized trials to come up with validated treatments, which are vital.
Yeah. I mean, when we think about lung COVID, there's the you know, I still have fatigue, or I have brain fog, or I've pots or something like that. And then there's mostly Dr. Ali who had on a several months ago studies looking at the long term risk of heart attack strokes, dementia, kidney disease, diabetes. I mean, I have to say they're scary as hell do you believe them? Do you believe that over in this population, we're going to see increasing rates of chronic non infectious chronic diseases across pretty high prevalence in the population with because of COVID infections?
That's a great question. Certainly the word as you mentioned, Ziad has been very troubling, but it's been reinforced by many other studies. Now we have both Korea and the US with significant increases of heart attack and stroke in the later latter follow up phase in excess of controls. Another one coming out later this week, the same. So the cardiovascular risk appear to be now replicated independently multiple. And then the other ones that you've mentioned, whether it's neurologic, kidney, you know, all the basically all the systems you know, reproductive there have concerns as well. So, I think we have to acknowledge that this is an area that only time will tell careful studies are obviously vital. But, you know, there's a lot of concerns about multi system involvement, appearing late and as you know, Bob, Parkinson's disease as a risk from the original 1918 pandemic didn't crop up for many years. And post polio syndrome didn't show its self for at least 10 or 15 years and even occurs in people decades later. So, there are things we still might not know, even though you know, the things that the science we have seen with respect to, you know, cognition with respect to, you know, diabetes, type two, especially cardiovascular, they're all points of concern. Yeah. Yeah. i
When people ask me, you know, how concerned Aren't you concerned about long term effects of the vaccine? I say, I'm 100 times more concerned about long term effects of COVID that we fold that we don't fully appreciate it. In terms of the literature on long COVID, are you convinced that vaccination helps, are you convinced the Paxil that helps?
I think there's evidence for both. Another really good study, in this week's British Medical Journal got to get to the vaccine protection. If you look at all the studies, there's ample evidence now that there's some help whether it's to lessen the severity of symptoms, the duration of symptoms. So or prevent long COVID. So I think the only question is, is it 20 Some percent or 40%, but there's some reduction, and the same is for PAX COVID is about a 25 or so percent reduction of incidence of lung COVID, which I think is one of the main reasons to take back some of it early to inactivate. The virus to stop it replication that is. So yeah, it looks good. And that's why I'm very surprised. It hasn't been a Paxil, randomized trial at scale. And people who use their suspicion that they have persistent virus. But yeah, both of those look like they're helping to prevent long COVID. It's the treatment side that we've got problems with, because vaccines don't seem to help getting a vaccine after you had long COVID. There's not much data to show that it helps. And there's no real treatment yet that's been identified as clearly being beneficial.
Yeah, yeah. So when you say there hasn't, you're surprised that there hasn't been a randomized trial randomized trial in treating long COVID not preventing.
Yeah, exactly. Yeah. Because there's a lot of evidence in some people have persistent virus, whether it's remnants of the virus or reservoirs of virus, and you think this chance to passively could kick in and do something or another potent oral, antiviral pill?
A few more questions on COVID, then I want to toggle to technology. You have done an immense amount of good in combating misinformation, or at least putting out good information. What are your take home messages after three years about the information ecosystem? And how well or poorly it's working and how we're going to deal with this problem going forward?
Yeah, this is a biggie. So all for almost every day, for three years, I've put out some information regarding the pandemic. I never thought I'd get into a situation. It's not my field. Yep. And I thought, you know, I tried to help. But it's gotten pretty hard to continue. Because the main platform of Twitter has really deteriorated. The reach isn't there, whether it's shadow banning, or whether it's just because there's so few employees at Twitter, I don't know what's going on. But you know, it's profoundly reduced. And the toxicity is gone to new levels, like I've never seen in we've seen a lot in the three years. So the idea that you could migrate to Mastodon or post those don't have legs. So the reach isn't there, even though the people are much more cordial overall. And so we have a problem, we've lost our main platform, and I've, I'm quickly losing my motivation to try to help on that matter. And I do like to communicate stuff that I read that I think is interesting, or, you know, weigh in, and I, you know, I've enjoyed the substack The problem was, I can't even open up for comments because of the toxicity migrates over from, you know, into that. So, I did that in the beginning. And that was over within hours, because, you know, the mob came on board, and they are organized. I mean, they set each other off. It's, it's highly orchestrated, and it's a sad state that we're in because we never combated misinformation disinformation from the get go. And it just has grown. It's funded in a lot of different ways. And we're we haven't anteed up and it's just it's a it's a dreadful situation right now. I don't know what you think.
Yeah, I mean, I sort of I hope you won't stop because I think you do a lot of good and You know, I tried to do the same and it feels like if we give up, it's, you know, you're letting them when I ignore the comments, I just assume it's just the cost of having a platform that they can then hijack and do what I can. And you know, enough people tell me it's valuable. And I hope they tell you to because what you do is extraordinarily valuable. But it is, it's hard to figure out how you get around it because everybody's choosing their own information ecosystem. And there's clearly a motivation for misinformation that some of its financial, some of it, I don't understand why you would want people to, to encourage people not to do something that can help save their lives. I don't get it. But there you are. But I think the biggest risk is that someone like us put the energy into educating people will say, this is just not worth it anymore.
Yeah, I'm getting close. Sure, yeah.
I'll try to buck you up. And maybe we can help each other.
Well, you've done a phenomenal job, too. And you've reduced your participation of lead averaging a
little bit, but I not so much, because it's toxic more. I just don't I don't want to do it, just to do it. I feel like when there's and you, I don't do what you do, which is really annotate the literature. I do, you know, here's how it's changing how I'm thinking. And it just doesn't change this as quickly as it used to. It's pretty stable now, for the reasons you articulated.
Right? Yeah. I just would say before the pandemic, it was a really great platform to share science and medical stuff. And your net benefit was, you know, palpable. And now, I mean, you know, even if you put stuff that's not related to COVID, you know, there's the, the people response is just, you know, it's ad hominem. It's, it's, and I don't look at the comments, either. It's just a den people who I do know, sent me to comments, right. I hope you're
doing okay. Wow. Yeah. Yeah, no, I've learned it early on my my younger son had COVID. I tweeted his case and with His permission, and at one point, when we were debating, like how long you should stay in isolation, it was day seven, the CDC said, He's okay, you can stop isolating, but it was his rapid test was floridly. Positive. And I said, I, you know, I love him to pieces, but I wouldn't want him to hug me. And people, what a crummy parent you are, that you wouldn't want to hug your child. You know, first of all, he's 28 years old. Second of all, he has a potentially fatal disease. It's like, really, that's that's where this is. So a little wild. But one or two more questions about COVID. I want to move on one of our listeners as immunocompromised people, would you recommend a six booster now?
Yes, I think you know if they can handle it, that I know people who've done that, and you can get it through the drugstores, a check immunocompromised. So yeah, I think that would be prudent. Okay. And just supporting not just mildly in compromise. I mean, that spectrum more immunocompromised, the more it's important to get it?
And is that because IV shell doesn't work? If Abby shell still worked? Would you be think differently about that?
Yeah, no,
I think if we had a new shell, that would be a winner. It's lost its efficacy, we have nothing else to offer. So yeah, that's, I think that's the main reason. Okay. And you've told us you haven't had COVID. And the prevalence is pretty low in San Diego and all that. So walk us through how you live your life in terms of masking in terms of indoor dining, what you do on airplanes, that kind of stuff?
Yeah, well, I will use a mask a KN 95 on planes and in airports, I have gotten back to traveling to some extent. So yeah, and you know, I don't go off into indoor restaurants but have more in recent months, weeks, he's been you know, very careful about the setting and I prefer outdoor and shoot for that, except when it is impossible. So I try to leave if I go to the grocery store, which I do, not infrequently, you know, I'll wear a mask. So just, you know, what are the kinds of settings where you just don't know I wouldn't want to be in a crowd, you know, concerts and things that you know, ballgames at the moment indoors. Just because you know, the there was a recent CDC study that showed on where they got the planes they looked at the wastewater on the plains Yeah. Where the from the plane at present had people on the planet COVID active COVID So you know, it's out there, it's still certain the lower it gets the better. But you know, I especially at this point, when I know I'm less protection because of being six months out from a booster, I just tried to be a little more on the cautious side, and you will
in an indoor restaurant in on occasions if that feels like the right thing to do. You shouldn't be prefer outdoor If it's the weather's good, and all that kind of stuff. Yeah, exactly. Yeah. You know, I've said for months, you know, the asymptomatic test positivity rate at UCSF, which we're not testing everybody anymore. But still testing a pretty good portion is still two to 3%. So if you do the math in a plane with 100 people, it's almost 100 guarantee that there's someone on the plane who has it, and feels fine. Yeah, so the wastewater doesn't surprise me at all.
Exactly.
We have a couple of the q&a say, Eric, please don't stop on Twitter, exclamation point. And so just you should know, you have a ton of fans out there. And the and you know, it's hard not to pay attention to all the the other stuff that comes in, but there's, I think, a silent majority that hopes you'll keep doing it. So I hope you do. Let's, let's switch over to digital. So I assume you've had a chance to play with chat GPT and the new AI when you first did it, what was it? And you've written about AI and the future of healthcare and precision medicine and all that. So when you when you first use it, what was your What was your? What were your thoughts?
Yeah, this is, to me, a very exciting and daunting as well, of course, it's hard to get one without the other. But it's my main area of research interest. And, you know, I'm all over it, that is the whole idea of getting self supervised, basically unsupervised learning. And that's what, you know, Chet GPT brought us with multimodal going from text to image to speech, you know, across the different domains. And it GPUs on Super steroid doses, as what we had in the deep learning phase. So, you know, it's very, yeah, and playing with it, you get an impression about its potency, and how it has, what we've been missing in medicine, you know, we had all these ideas about synthetic notes and keyboard liberation, and not having to deal with pre authorization, you know, automated up notes, automated bedside, you know, round notes, and we can do all that now. I mean, that's amazing. We can do all that stuff. We just need to get a chat GPT or Lowery's language model, generative AI foundation about or whatever term you like, we just have to get that to be pre trained medically. Right now, you know, what you're what we're working with, if you get on to being which is a bit a little bit more sophisticated than Chad GPT. And it'll be even more. So Chad CPT for that's coming soon. You you see the power, but it's not medically framed. Once it is, it's going to be a big, I think net positive shakeup. But it still requires humans in the loop. You know, you still have to, you know, review things. But it's really exciting. And by the way, I do want to compliment you on your team. And that grand rounds a week or two ago on this topic. Erin, and Sarah and the nominal grand rounds that if anybody's interested in a topic, it's worth spending that hour to listen, it was excellent, really well done.
Thank you. It's such a incredibly fascinating and rich, rich topic. I love sort of having just listened to a New York Times story on Elon Musk and self driving cars, you say it's still important to have humans in the loop to check in be sure. Do you think that'll happen or you think as this AI gets better, we all kind of collectively fall asleep at the wheel?
Well, you know, the level five self driving car will never happen. So you know Elon Musk, who is not a man of great integrity, or character who keeps saying it will is wrong, because that means it has to work in all conditions, and it can't. And so he's been hyping that up. And that's part of the problem with AI. It's had hype for decades. And that's why at one point went to, you know, a very serious winter. At this point, it's it's the issues are, we have to balance its expectations. You know, you can't just have all these things that are mentioned about keyboard liberation, where you see a patient in clinic, and then not just the note generated, but the future appointments, the labs or whatever tests that are being ordered. prescriptions are done, you know, you can't do that without surveillance without editing and, you know, getting it all validated. But, and then the other thing is, is UCSF or a different institution is going to roll on their own or we're going to have some that are, you know, generally applicable on a wide scale basis. Right. These are some imminent questions, but there isn't any question in my mind now that we have solved the missing link of how to go from the data to generating you know, the components, the pieces that we need in medicine. Because right now, if you were to say, what would be the immediate thing that could fix a lot of the despair of clinicians, the data clerk functions that are so burdensome to track from care of patients, it would be this. And I think it's imminent and more imminent than than ever before, that we'll start to see it. But the human in the loop, you know, just because it generates all those things, you can't not look at it. You know,
what, I guess I'm the IT that sounds right. And yet, that's what I tell people periodically is like, I don't remember my wife's cell phone number anymore. Because it's in my phone. And so I've turned that part of my brain off. If I ever lose my phone, I may never see her again. Because, you know, you just get your brain gets lazy you. And so in some ways, it's like the self driving car, you wonder whether it reaches a point that it's so good, that yes, the human should be in the loop. But the humans heart can pay attention when it's right 98% of the time. So it's tricky to get that balance, right.
Yeah. And there's other motivations, too, you know, this started with the hit your Ristic models of ECG interpretation. And then you know, you're getting reimbursed by overreaching ECGs, and you're busy, and you just sign off on them. And it isn't just because you can't read ECG, but you have other reasons to not have the oversight, you should know there are concerns there that, you know, atrophy of surveillance or, or lack of incentives, or whatever. So yeah, that's all part of the human loop, the real human intent, human in the loop, right.
And you've written, of course, about how at least the hope in theory is that it frees us up from these mundane data clerk kinds of tasks and opens up time for humanism. And that's a perfectly reasonable point of view. And the other reasonable point of view is that time will just get filled in with more tasks, do you think you're confident that the time will be protected and will be used on the things that humans are uniquely good at doing?
Right, so the theme of gift of time, which is what this can bring us. And it really wasn't realistic until we saw the large language models. But the gift, the gift of time isn't just on the clinician side, remember, it's also on the patient side, going to be capturing their data and having it algorithmically interpreted. And most of the common diagnoses will at least be screened. You know, before Doctor lists before consulting with a clinician, so there's a lot of drivers to get us the gift of time and get us into caring mode, which is, you know what Aaron 19 elaborated so nicely. I think that the issue that you're bringing up is the overlords. So we have these overlords. I can't speak now to the current ones at UCSF, but throughout this country, we have administrators who are, you know, financially driven for revenue. And they have no motivation to say you have more time with patients. In fact, their motivation is read more scan, see more patients read more slides, and on and on. So the only way we can we can buck that is to have a gravitas of the medical community standing up for patients and our profession. And we haven't seen that if we have no organization that is ready for that. You know, basically what I wrote about in the New Yorker, we really need to unite and not let what you're describing happen, because by default, that is, you know, the financially big business of medicine and healthcare. That's what would happen. Yeah.
So Nora, girl, she blogger says, quote, I have unique total stories, but we're not going to give her a chance to to, to talk about them. But then she says, seriously, please tell me how I will not be put out of business with AI. So what do you tell a young doctor, as they look at these models, and just say, am I gonna have a job? I just spent $200,000 on my education and all of that. And you know, I'd love to have a career, right, like Eric had, but 30 years from now, what's going to be the job of a doctor?
Well, I wish I could go back because the AI world will ultimately get us to be far more accurate and making diagnoses. It'll give more autonomy to patients, those who want it and then be spending the time that's needed for serious matters. And I think that getting back to the initial mission of why we went into medicine, which was caring for patients, that they know we have presence they can trust They have that that patient doctor relationship, that bond for humanity of medicine, that it can be restored. I think it's the most exciting time ever in medicine going forward. As we, you know, get over these, these gaps, from theory to practice. But no, no, this is not about replacing anyone. It's about what medicine, the essential reality of medicine, which human human connect, and, and letting a lot of the other stuff get done or facilitated, improved. I mean, you know, how long it takes to review a complex chart of a patient? Imagine that's all done for you accurate? And for you even, you know, start to look at the thing. So time is what is the AI gift to us, so that we can get back to real medicine? I think.
Yeah, I hope so. If we get it right, that would be terrific. Do you think the earliest use cases are, in some ways, sort of taking the encounter? And just alright, I don't have to document it. Or I needed to write a letter to the insurance company to get permission for the MRI, it kind of knows that I that's what I do. And it does it in my style? Or do you think it's more at the level of it's helping me make the diagnosis, it's looking at what's going on and saying patients like this turned out to have lupus and make sure you send these tests? Is it sort of on the operational kind of get stuff done boring part? Or is it on the more essential doctoring part that where do you think's gonna have the biggest impact?
Well, I think both will be major fronts. But the one about operational maybe the one that is of more critical need, because doctors don't perceive their diagnosis issues right now very well, even though we know there's more than 20 million a year of serious diagnostic errors. So they don't perceive it, because it's, you know, unusual for them, but they do perceive the operational issues, and
it's either unusual, or they never find out that they got it wrong.
Right, exactly. So, you know, but cumulatively, once we get over the operational and concurrently, we'll address the accuracy. And I think that's going to be really another big big step. But you know, they're not mutually exclusive. But what's the crisis right now? I think we can address that pretty effectively.
Yeah. You've written a lot about the democratization of care. And in some ways, one of the things that technology always does is democratized stuff into sort of means you don't have to go see your accountant or the travel agent or the banker, this, you know, the technology helps you do that. Well, how do you see the balance of that I, you know, from having read you, I know you think it's net good, you have any worries about that, that patients will have some ways false reassurance that they have made their own diagnosis, or they're managing their own care, and it's sometimes maybe they should be seeing a doctor, how do we get that balance? Right?
Well, I think we, we know what we have right now, which is, you know, Google, and other websites. And that doesn't work very well. Because basically, people looked at things up, and whatever they read about they have that disease, there's no specificity about their actual objective data. So here, you know, we have advanced searches, which is going to be happening in healthcare, where that person's data specifically, is used to help, you know, give a differential of what is happening. So I think that it's going to go to a better level, it doesn't mean we're not going to have cyber contracts. And we're not going to have you know, people misled, but it can't get much worse than it is now. So I'm actually, you know, thinking, this is the reboot of search, which is all part of this kind of revolution will rather than trying to look at all the different hits on multiple pages, where it's really laid out, in a way you want to be is really going to be a plus, I think,
yeah. Do you think it'd make health care less expensive? Someone who's who we thought was talking about quality and patient experience democratization, and these days, we talked some about equity, which is great. But ultimately, in some ways, the biggest crisis is that it's 20% of our GDP and bankrupting governments and, and businesses and individual people. So do you think that is that part of what you think is likely to happen?
Well, that's the trend if we don't again, try to arrested or sidestep it. So you know, they're the biggest potential which is going to happen. And I wrote a sub stack on it hospital at home. So of the $4 trillion, we spend a year. The biggest chunk is hospitals. And of course, we have perverse incentives in this country because the American Hospital Association is one of the largest lobbying groups and the last thing they want is to have hospital at home and there is no reimbursement except temporarily because of the pandemic for hospital home. But now with the AI, with all the different capturing of data, and this is not with clinicians in that in the home, this is only going to the home if there's algorithmic prediction that there's trouble afoot. So, you know, I think this is got, if we can get hospitals, except for, you know, the ICUs, and emergency rooms and operating rooms, but if we get regular hospital rooms, and we keep people in their home, right, if they have a home, that we can take out a huge part of our annual costs in healthcare, so that what that is, that's further away, I acknowledge that it's not going to happen quickly. But if we were serious about cutting costs, we would get on it, we would do the validation studies. And we would, you know, we might just do it for heart failure, or pneumonia or certain certain things and then broaden it to all comers.
Maybe get it connected to the end, as you have, as you've been studying the digital transformation of medicine over the last 10 years. What are what are the lessons that you take away that you think will help inform what the next 10 years are like in terms of you know, is it harder than it looks or easier that looks or obstacles you wouldn't have expected have kind of emerged themselves. But how do you how do you see this pretty bumpy path so far?
Well, I'm an old dog. So I know it takes a lot longer. You would ever think it would be fathomable? I mean, when I saw the light of digital medicine, you know, with the iPhone and oh seven and how little we've actually progressed in digital medicine theory are 16 years later. You know, it's sobering. There's lots of obstacles we get in the way of ourselves, we are our own worst enemies in medicine, of progress. And instead of trying to get validation, we just let stuff sit and kind of let the potential get enact not not actualized. So I hope that we can move things forward. I mean, there's a lot of exciting things happening right now, not just in AI, obviously, with genome editing of incurable diseases, but we don't know how we're going to pay for it. You know, there's, there's so many things that on life science and medicine that are exciting, that hopefully won't take, you know, 1517 years to actually get get into medical practice routinely. I mean, to me that being so impatient, that's the hardest thing is that we eventually get there, but it takes so darn long and unnecessary lost opportunity.
Do you think precision, you think we're, I mean, we've been talking about precision medicine kind of being around the corner for 20 years or 30 years. And it feels sort of real ish, in certain parts of medicine, like, like in oncology, and maybe a little bit in cardiology, do you feel like we're at the cusp is, is that things have changed enough technologically and or politically, culturally, economically? That that is going to be it's going to be very real in the next 10 years?
Well, it couldn't be more real today, if we had pharmacogenetic tests that can be done cheaply, and, you know, immediately, or over 150 drugs that have a genomic label, but you've touched on, you know, in oncology, there's no question about matching up mutations, with drugs, although, you know, it isn't always saving lives and in curing cancer, it's making some improvement. I think that genome editing if that isn't precision medicine, and nothing is right. And we're already seeing cures of diseases that you know, we're not there was nothing to offer a book so yeah, it's unfortunately it's not a light switch. It comes slowly in neonatal ICUs a baby comes in today you don't know what the diagnosis you do genome sequencing you make the molecular diagnosis and give a definitive treatment you prevent the baby's death or or end organ damage that's irrevocable that's precision. So we're going to see it it's just not what you making diagnoses molecular diagnosis of rare unknown diseases eventually we'll get there eventually the accuracy and medicine not so much the precision will be promoted and we won't be using this stupid term anymore about precision will just you know get to a new level of medicine that is just the medicine Yeah, yeah.
We won't be talking about digital medicine either. It'll be medicine well I when I tell I tell you know the house staff periodically you know, about what it was like good you know, paper charts and go into the VA down to the lab and going through the shoebox looking for your lab slips. They of course, you know, 10,000 years old but
that we do our own gram stains and our own real the patient in the mission to get their X rays. Oh, yeah, no, they think it's just crazy stuff, right? The days
of the giants. Eric, thank you so much. And we at UCSF are incredibly proud of all the things you've done and hope you keep it up in terms of getting the word out and getting through that it's really been important and really transformative for those of us following along. So thank you for everything
deeply. Appreciate it. Keep up the great stuff at UCSF.
