This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.drvinayprasad.com/subscribe
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
Transcript
From the show. Tap a time to play from there.
SPEAKER_00i recently read the book how to rule the world by stanford graduate theo baker a budding journalist and it describes his undergraduate experience at stanford which culminated in taking down the president of the university mark tessier levine mtl they call him i'm gonna give a brief summary of the book and i'm gonna give a brief summary of my thoughts on this issue of course the book is sort of a coming of age story of being in college where a young man who has all the potential to be a rich investor in silicon valley or be a budding computer scientist decides to pursue journalism which is the trait
SPEAKER_00of his parents it's a riveting story in that but what makes it even more interesting is of course in the course of his undergraduate studies he writes a series of articles that culminate in the resignation of mtl the president of stanford this is a gentleman who previously was the head of rockefeller who's published many provocative papers many important scientific papers in his career on alzheimer's and other neurologic conditions when he worked at ucsf when he worked at stanford when he worked at genentech where he was a vice president of clinical science, and then we went to Rockefeller in at Stanford.
SPEAKER_00And what Theo Baker uncovers, which was kind of hiding in plain sight, was that a number of these papers on PubPier and the sort of secondary and tertiary servers looking at published literature had identified that they were images in some of the papers that looked like they were, for lack of a better term, photoshopped duplicated, modified, photoshoped, something like that. and the more they dig on those papers the more they find that there are problems with those papers that they didn't replicate in other laboratories
SPEAKER_00and perhaps there was some element of photoshop that went on Who did the photoshopping? I personally don't think it was MTL, but i do think like many labs and this is a phenomenon across many institutions there are some junior people under the tremendous pressure of generating results who resort to Photoshopping. And PI might just run with that because its convenient and natural and easy and helps their career or the pi might be purist and scrutinize everything in which case they may never publish papers and get booted out of the Academy altogether. in this case it looked like mtl had a pattern of supervising papers
SPEAKER_00across multiple institutions with this problem and that became a problem for him at stanford because people allege that science is pure it's about trust and if we don't have trust in you you can't be our president here are some thoughts i had after i read this book so my first point is i i'm persuaded by theo baker mtl has a problematic research history papers at different institutions with different co-authors all have this pattern at a minimum he was slow to correct the record at most he perhaps was somewhat complicit in this it is a problem for him from a science standpoint number two i believe people
SPEAKER_00who commit repeated scientific infractions should face some punishment for that number three i think people should be investigated proportionate to the impact of their claims so can't just pick on college presidents because they happen to be presidents like claudine gay plagiarizing some of her papers and she's the president of harvard and mark tessier levine and stanford you should pick on people who make big claims in the literature in nature science cell jbo big claims and they should be scrutinized equally well that's not what's happening here
SPEAKER_00journalism focuses just on a few anecdotal stories and when they find something wrong they write a series of articles and when they get one guy fired they say congratulations we solve the problem but they're not solving this problem in a systematic fashion and you really need to address it in a systematic way number four i think one of the things that led to the downfall of this individual and so many individuals is that they provide the raw imaging for certain types of science view to a western blot you provide the raw imaging that's the raw data of
SPEAKER_00the experiment you don't do that in epidemiology people are mostly not uploading their data set and code you don't do that in clinical trials of course that data is a secret they like to say you don't do that in so many other types of science where you just show a figure of the results and not the results themselves mtl like so many scientists are getting crucified because they're giving you the raw information and we say wow the problem is right here we don't realize the problem is everywhere if you made everyone give you the raw information
SPEAKER_00of epidemiology papers i suspect sixty percent seventy percent or eighty percent would be found to be false fraudulent or worse or maybe slightly better but probably worse there'll be something really problematic with this data in the course of this channel in the course of my Substack for years i have taken very prominent papers that are covered in the new york times time magazine all the best science publications and i show that even though everyone is showing how great this study is oh boy the shingle shots prevents dementia i very quickly can find half
SPEAKER_00a dozen things wrong with this study that make you doubt the conclusion If I had access to the raw data, I would be able to obliterate far more studies. I mean, I could redo the analyses. I'm sure I'll find problems with so many things. So I think what's getting lost in this narrative In Baker's excellent book What's getting loss is that MTL Is not an exception to the rule, a guy who broke the rule.
SPEAKER_00the book falls mtl for being too ambitious they have some quote where it says something like somebody worked in his lab said he really liked to spend time with people who are generating results and he kind of ignored the people in his lab who weren't generating results he was pushing people to find the answer to alzheimer's and other riddles of biology he was very ambitious he wanted to rise in academics i hate to break it to the author that's everybody okay if you're gonna fault the guy for being ambitious you got to go to these universities and literally you can fault every single person who works there you don't work at stanford or harvard or even the university of kansas without ambition to get to those positions you have
SPEAKER_00to be ambitious every single lab pi is showing favoritism to the people who work hard and generate results that are provocative and not to those who aren't and you know it's not a super fair world you should show favoritism based on how hard people work but somehow it's very difficult to gauge how hard people work and people use results as a surrogate for that so i don't know what to say i think that's kind of distracting from the book ironically being a university president and being a lab researcher have nothing to do with each other lab researchers are the kinds
SPEAKER_00of people you want to be honest faithful and pure university presidents just need to shake hands and raise some and so mark tessier levine is being fired from the president job not a lab job he's being fired as president so he can spend more time doing research in fact in many ways the punishment doesn't fit the crime it probably would have been better to have him as president it's quite a good fundraiser and probably worse to have him in the lab a person like mtl will have a lot of enemies the book doesn't talk about this but anyone who rises to the top of these entrenched political organizations will have a lot of enemies
SPEAKER_00and those enemies will use anything to bring him down mr baker describes that he received sort of an anonymous complaint one year to say you should look into the president but we really do want to know who sent the anonymous complaint was his laboratory rival and if we get into a situation where of this ocean of scientific discovery the only people we investigate are those in whom a rival triggers a journalist to look into it that's not a great system that looks like a very sort of chaotic and cruel and vindictive system and not a fair and systematic way
SPEAKER_00of solving the science problem journalism of course is not really fit for purpose for evaluating science journalists don't know science the author of the book called though excellent he had to rely on other people to tell him if this is a problem or not a problem and journalists are love to tell stories about individual people as bad actors this is a sort of a a greek tragedy of an individual but that's not how the world works people respond to incentives and there's an incentive in science to generate discovery to generate claims of discovery and not to replicate those claims and that's
SPEAKER_00the problem it's the system it's not the individuals so i think ironically journalism is not the cure to the science problem the whole story of mtl reminds me of jose baselga baselga some of you may remember was physician in chief at memorial sloan kettering which is one of the most important oncology roles in america and some people may think that prasad and baselga used to have a bit of a rivalry there is some truth to that we disagreed about some fundamental claims being made and the level of evidence for those claims and we had a public debate on that at acr but behind the scenes
SPEAKER_00i must say i always liked jose he's a very funny guy and he is very smart and he is very driven and i always admire those traits in someone even if we disagreed about a bunch of things and every time we met face to face he made me laugh because he says very provocative things and he says what he thinks and in this line of work i would give a million dollars to have more leaders who say what they think and don't hide behind a cloak of being quiet about every single issue so I actually liked Jose as a person I noticed many years ago in the New York Times somebody tipped off
SPEAKER_00a journalist Charlie Orenstein to the fact that Baselga had published some papers that didn't disclose all of the money he had been getting from biopharmaceutical firms he disclosed in some but not others there was inadequate disclosure in some papers he messed up the standard was to disclose he didn't disclose stop for a minute does disclosure actually solve the problem of conflict of interest actually no it doesn't it actually more likely to lead people to falsely believe the person is honest when there's something to be scrutinized doesn't solve the problem at all recusal would solve the problem but nobody in this field wants to do that
SPEAKER_00so we have this token effort called disclosure and he failed to have universal disclosure was he the only person who failed probably not there is probably an ocean of doctors who failed to disclose but he was in a prominent position and someone who doesn't like him tipped off the reporter and sure enough they investigated there's a series of stories he resigned from memorial sloan kettering and then he was punished by going to astrazeneca where he was paid probably five or six times as much money as all punishments go and then that was his fate unfortunately a few years later he would pass away
SPEAKER_00from kreuzfeld jacob disease a sad story my point here is simple you can have rules and if you want to enforce them you should enforce them equally you can't just pick one Jose Baselga make a scapegoat out of him pick one MTL make a scapegoat and then say problem solved now a decade later in oncology is conflict of interest solved spoiler alert is worse than it's ever been And when I talk about conflict, there's one particular type of conflict that is problematic where it is the doctor who prescribes or advocates for some therapies in
SPEAKER_00an off -label manner receiving money from the company that only profits from selling those therapies. This unidirectional profit mode which I've written about in some publications. That's the real core problem. Regarding MTL, completely separate from his research foibles, he made a decision… the covid pandemic that i think is far more concerning for a president scott atlas you may remember a radiologist at stanford became donald trump's advisor in covid 19 policy and atlas had these crazy ideas that lockdown did more harm than good
SPEAKER_00that we shouldn't close schools that cloth masking was mostly empty virtue signaling these positions are miss actually completely correct they were completely correct at the time they're completely correct now he was punished for having completely correct views the stanford senate censured scott atlas and the university president said he was a purveyor of misinformation and stanford has to get away from that so if you had to ask me why should mtl be resigning as stanford president should it be his possibly photoshopped western blots
SPEAKER_00or should it be that he didn't defend academic freedom when it really mattered it should be the latter that's why he should be resigning but he's resigning for the wrong reason It's much more telling that he didn't defend Scott Atlas, and in fact cast stones at Scott Atlas. You know it's easy to defend academic freedom when people say things that aren't that objectionable in times that don't matter that much It's very hard and very telling to defend it when it actually does matter when things really do depend on having that freedom And Stanford failed He wasn't the only one Of course Bob Harrington
SPEAKER_00who was then chair now dean Failed terribly with Iran Ben David and Jay Bhattacharya in trying to quiet those individuals from saying that lockdown probably has more harms than good, a position that now we realize is probably completely correct. Yeah, it's a shame that we didn't allow these people to make their case more forcefully at universities. the king is dead long live the king theranos is gone long live theranos no other companies making false claims i say this in my Substack essay on this topic because my point is that even
SPEAKER_00though mtl is gone we over and over again see many research articles with likely completely fabricated and false claims i talk about them endlessly on my Substack you should subscribe at the same time despite all these forces against us science is still making progress there are still discoveries and cures and proponents say as long as there are discoveries and cures there's no problem in the workshop there's nothing wrong back there we're still curing diseases it's true we're still curing diseases and it's true that mtl still has real discoveries i mean some of his work led to denali
SPEAKER_00and denali has a successful drug approval in hunters he is still making progress but this is like one day your kids say hey i want to make you pancakes and you say go ahead knock yourself out go to the kitchen they come back and they've made you one pancake and you go in later and you find out two dozen eggs are broken they've used all the flour in the house pounds of flour there's a mess all over the place you're gonna say i really appreciate that you made that one pancake but in the future we have to learn to make pancakes a little more efficiently and that's true in science they're making pancakes they're making the product
SPEAKER_00but they're making it so inefficiently and there's so much graft and theft and fabrication and false results along the way and people are padding careers on nothing but false results and occasionally giving you a pancake and telling you you should be grateful for it that's a problem in any normal industry where there are market pressures you wouldn't get away with this level of bloat and waste unfortunately in this case it's mostly subsidized by the us government so you can get away with 99 waste you can't get away with 100 waste but 99
SPEAKER_00and we may be very high in waste we may be in the 80s or 90s percent we don't know there's no effort to ever quantify what percent of science is true or useful people like john yonides and others have estimated that it's very likely that most science is neither true nor useful but it's never been properly quantified and the way i would quantify it would be You know, at random you pick 1,000 scientific papers randomly picked from the literature. Maybe you waded towards papers making more provocative claims perhaps or maybe not and you try to replicate those papers just straight up 1, 000 papers and you'd find this has been
SPEAKER_00done piecemeal in bits of this replication rates are abysmal 40% 50% in economics and biologic science and cancer research etc really poor replication rates and if you looked at the literature more broadly, in the ways I've described it might be even lower than that. and if you can't replicate it really raises the question of what actually happened in the first experiment i know some people like to say well you know it was the fact that the wind was blowing from the southeast that day that led it to result that way well if your study is detecting the wind blowing from the southeast you're not learning essential truths
SPEAKER_00of biology are you you're learning idiosyncratic results that occur when the wind is blowing from the southeast and that's not the science we want to fund so there are a lot more points in my Substack article but this is the gist of what I want to talk about it's a great book it's a fast read it reminds me of John Kerry Rue's Bad Blood which I read on a single airplane flight kind of to speed through the book I enjoyed reading the book I enjoyed learning more about the MTL saga I think that the lesson that's being taken away one guy Greek tragedy kind of story it's not the right lesson the lesson should be a broader lesson about science in general
SPEAKER_00and that wasn't fully captured in the book which is what i'm trying to do in capturing the video so if you like this video subscribe to this channel hit the like button recommend it to a friend and i'll be back with more videos
Live episode from Youtube This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.drvinayprasad.com/subscribe
Elizabeth Holmes/ AI researchers start smoking/ Moderna vax by Plenary Session This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit...
Pope has cancer; Phase 1 trials for AI; Can AI hack cancer trials__1 by Plenary Session This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit...