Bad science is a big f'ing problem. Medical research edition. Roundtable w/ Reese Richardson PhD & Rohin Francis MD PhD

0:10Rohin Francis: Welcome back to the ECO-L podcast. This is a show about the science of cosmetics. My name's Jen and I'll be your host. In case you weren't aware, bad science is a really big problem driven by some of the very skewed incentives within science and publishing right now. I've covered this topic from many angles over the years. For this episode, I wanted to get into the angle that's been on my radar a lot lately, which is bad science coming from the medicine side of things. What's been happening? Why is it happening? What are the repercussions? And what should you know? To help me do the topic justice, this episode featured a roundtable discussion with research integrity researcher Dr.

0:53Rohin Francis: Reese Richardson and cardiologist Dr. Rowan Francis, a.k.a. MedLife Crisis. Before the episode, thank you so much to my Patreon supporters for supporting my SCICOM work, including this podcast. Thanks for helping me continue to create content like this and invest the time into going down these rabbit holes I've been able to on the show. If you'd like to support my work on Patreon and it's accessible, you can find me at patreon.com. Other great ways to support this podcast is by subscribing to it on whatever platform you listen to it. Please review the show. Your reviews help more people see the podcast, which is also a huge help in keeping this podcast going.

1:36Rohin Francis: With that, here's the episode.

1:42Rohin Francis: so this is going to be a conversation about information integrity it was inspired by a was it a tweet or a thread the post was new data newly minted doctors applying for residency in the u.s now have between four to ten published articles abstracts by the time they graduate medical schools and apply for residency. This environment breeds misconduct. I didn't actually say who I'm talking about, and I realized this is an audio recording, so I'm talking to Dr. Reese Richardson, the research researcher. My husband, James, came up with that after our last conversation, and I felt really bad for not using that description.

2:31Reese Richardson: You got time to workshop it before I'm back here again.

2:34Rohin Francis: And I'll also say that to help broaden this conversation past to researchers' research integrity, I have Dr. Rowan Francis. You may know him by his YouTube channel, MedLife Crisis. And he has a very unique perspective to bring to this discussion in that he has both an MD and PhD, which to me as somebody going through a PhD is honestly unbelievable. And my name begins with R. I need to workshop this once we come to the title of this conversation. I'm going to step back a bit because I feel like a lot of people don't know, unless you've listened to my podcast a whole bunch, maybe you're new here, don't know about the issue of bad science and what that's looked like over the years.

3:24Rohin Francis: So I guess for some background, Reese, what is the problem right now with bad science? or what are some of the problems?

3:34Reese Richardson: Right. Where does one even start? So basically, we'll talk specifically about fraud, even though general reproducibility in science has been a hot button topic in recent years. Usually fraud in scientific publications is considered something that is very rare. It doesn't happen very frequently. And it's up to the actions of individuals that decide that, oh, I'm going fudge the numbers for this, or I'm going to make this thing look different, or I'm going to pretend that I did this experiment when I didn't. This can be a helpful way to make things move much faster if you're a scientist, right? So experiments take a lot of time, they take a lot of effort, they take lab space, they take resources, you might have to hire people, and many people are without those resources.

4:25Reese Richardson: So this presents a convenient way to make that happen. However, in recent years, we have seen the emergence of services that basically offer to do research for you, or at least make you appear as if you've done research yourself. We typically call these services paper mills, and they've been a big focus of my research. And the idea is, is people need papers, scientific papers, to present to their employers or to potential employers to say, I have this many papers that have gotten this many citations, please hire me, please give me promotion, please give me tenure. These papers are entirely fabricated, and they sell authorship positions on them, again, to people that are desperate for them.

5:14Reese Richardson: This has buoyed the rate at which we see retractions of scientific publications happening in recent years. We estimate that the bulk majority of retractions that now take place are a part of batch retractions that ensnare multiple articles instead of one article where it looks like something is wrong. It's many articles while being retracted at once. And yeah, as far as we can tell, it's a growing industry, even though it operates largely in the shadows. And this touches many different professions. It's not just they're not just selling papers, They're also selling feces. They're also selling books.

5:52Reese Richardson: They're selling fake patents for people to put their name on, to claim as inventors, anything that could possibly be valued. And indeed, we're seeing new businesses crop up that are preying specifically on newly trained medical doctors. I hope that gives an overview of the sort of bad science that I study. But obviously, there are many different topics that we could touch on. My PhD advisor loved to cite the Anna Karenina principle that is like all happy families are alike. All unhappy families are unhappy in their own unique ways. The same goes for research going along, right? There are many different ways that misconduct and irreproducibility can rear their ugly heads.

6:35Rohin Francis: Rowan, coming at it from both the medical side, research side, but also the science content creator side. How has this taken shape from your point of view? You know, coming back to the original thread that inspired you to invite us to chat today, which was kind of talking about doctors in the US playing for residency, but it's a global phenomenon. You know, medicine is a field where I think it's just easy to look at as a succinct thing, but I'm sure this applies across all scientific endeavors. There's been this kind of general inflation of requirements. And I was talking to a friend whose kids have just gone to college in Philadelphia.

7:21Rohin Francis: And he was saying that even now applying to get into college, you're expected to have some kind of published journal article, which is absurd. And the American Heart Association, not long ago, a 14-year-old was presenting. And you could look at this and think, oh, that's great. What a sweet, overachieving moment. But it is actually indicative of this kind of ridiculous inflation that now, as your stat suggested, doctors coming out, just applying for residency, having multiple published papers. And how does this tie in with the topic of the discussion is because this is perfect material for the kinds of fraud that re-studies, but also just awful quality scientific papers being published.

8:04Rohin Francis: I plucked this stat out out of the air, but I often say kind of 80 % of scientific publications are not even worth reading. And indeed, that is supported by, there was some analysis, I can't remember what it was now, again, I'm plucking something out of the air, but you just take my word for it. I'm giving you an example of bad science right now. But I think something like 85 % of papers are cited once or zero times or something like that. There was a question once that somebody asked that, would this paper exist if it wasn't required for a job application? And I think that was such a good way of looking at it, that is it actually scientific curiosity that's led somebody to publish this?

8:43Rohin Francis: Is it some genuine interest? Or is it that they just have to tick a box on their CV or their job application? For a young person to have 10 scientific publications, I'm sorry, especially if it's not their full-time job, if they're a medical student or if they're a full-time doctor, those research papers are not going to be good quality. There's no two ways about it. Now, some of it is straight up fraud, like we've mentioned, but some is just predatory journals accepting really bad quality, you know, all the stuff that I'm sure your listeners will have heard of before, like publication bias and p-hacking and underpowered studies and all that kind of stuff.

9:18Rohin Francis: But, you know, there are so many different ways, as Rhys was alluding to, how you could produce bad quality stuff. In answer to your question about, you know, my different experiences, is I know that as a medical doctor, I was trained to appraise a paper, but, you know, in a kind of rudimentary way, like it's not a fundamental part of our education. And that's fine. You know, I think most doctors are not going into academia, they can understand the basics of data, but true scientific analysis, and really that deep understanding of whether a paper is good, whether the research is good, actually, genuine critical appraisal came much later when I went into research full-time myself for several years.

9:56Rohin Francis: Your question about science communication, I think that is sorely lacking. And some of the biggest, in fact, I'd say all the biggest science communicators do this on a regular basis. They just misinterpret or misunderstand data and often repeat things that they've read without really understanding the shortcomings of a particular study. So I'll stop talking now, but hopefully that kind of kicks us off. Before I flip back to Reese, I will highlight another commentary. This time it was on Twitter by Michael Barnett about a recent JAMA study. Over a decade, research abstracts, papers in residency applications have increased two to three times, which seems like a problem.

10:43Rohin Francis: so what could be the problem with this change for medicine and also just like for society and science what's wrong with this other than like oh there's bad science but like what's it doing I think you know from a professional's point of view as in a non-scientist from a from medical professional's point of view we want to give the message that academia is if you've got a scientific curiosity and passion, then you should pursue academia. It shouldn't be something that you're just doing as another tick box. And the problem is, is that in medicine, to be a good doctor is very hard to demonstrate, right?

11:27Rohin Francis: If patients really like you, and if you're a kind person and you listen to your patients, how do you record that? How do you show that in a job application? What can you show? Hey, I've got 100 different publications. And the same with other things that people exaggerate on their CV. So it means that people are not really learning the lessons we want for them to get from academia, to be better clinicians, or to have a genuine interest and become clinician scientists or full-time scientists. I think it's completely warped the average doctor's understanding or appreciation for science because, you know, so much of the stuff they're dealing with is kind of low quality and lousy.

12:10Rohin Francis: And secondly, I think we're in an era, I think the pandemic really showed us this, that a lot of this research is available to the public. And if the majority of it is poor quality, it just leads to more confusion. And this feeds into a never ending cycle of, you know, all the kind of misinformation and disinformation online. Sorry, I dumped in there, Rhys. I'll let you on through.

12:32Reese Richardson: No, that's okay. You make a really good point. So yeah, obviously this really warps the landscape for training as well, because it starts to reprioritize things that people are actually focusing on, maybe incentivizing things that we don't want people working on. But it also completely reshapes the information landscape. If you are a scientist or a clinician or a member of the general public that's trying to get information on something, it is much harder to actually get reliable information if that reliable information is swimming in a sea of bullshit, right? I'll give you one example of this that was actually published, I believe, just yesterday.

13:17Reese Richardson: There's this data service called TrinetX, and it's available in a lot of hospitals and medical training settings, and it collects a lot of demographic and treatment data about patients. And effectively, you can analyze this data any way you want. And indeed, the company that runs TrinetX makes a lot of money off of selling access to this service. One disadvantage of this is that you can write a publication on association between this outcome and that outcome or this drug and a particular surgical outcome like that. Over like the last five years, the number of publications publishing using TrinetX data has gone up from around 50 to around 3 ,000.

14:05Reese Richardson: You can use this data to find anything effectively. And indeed, we've seen a lot of a similar explosion in publications for publicly available data like the USNHANES nutritional health and something or national health and nutrition something survey from the United States, UK biobank data, etc. all using these publicly available data sets to make spurious correlations between things like how many noodles you eat a day and your likelihood for coronary heart disease. Many of the people that are listed on these are aspiring medical doctors or people in their last year of residency or what have you.

14:54Reese Richardson: So yeah, this is an investigation that was just published yesterday in science by Fred Jolving, who had previously written about the paper mill industry and its bleeding into the medical profession. That's my biggest concern is how the information landscape is being reshaped by the pressure to put something out there.

15:15Rohin Francis: I actually had prepared another example that's very similar, which was exactly the same thing that it was the National inpatient sample, which is, again, like a databank that you can access. And there was an analysis that found 85 % of papers using the NIS failed at least one, maybe more, of the methodological criteria that you would need to look at the data set. So essentially, people are using these vast data sets, as you say, either to just create spurious correlations. I think there is even a website where you can get genome-wide association studies with just any two variables, which is quite entertaining.

15:54Rohin Francis: They're using these data sets without knowing how, and they're just producing rubbish science. And the thing I was going to mention is also now in the era of AI, it's even easier to just generate this stuff. You can just plug in a whole massive data set and say, just pull out some and format it like a paper, done. Stick your name on top. It's very tempting when I say it out loud. If I can boost my H index up this way, then why not? But I do get it. I get the temptation. But as you say, it's just completely kind of ruining the integrity of published science.

16:27Reese Richardson: Yeah, actually, the science investigation quoted a new report that was released, well, in February by a scientist named Joshua Wang. And they tested, you know, asking different AI tools to interface with the TriNetX data. a bunch of these AI tools suggested using a controlling procedure that is not actually possible in TrinetX to account for things like collider bias or a mortal time bias, which can make these analyses go awry. But yeah, most of these studies are not accounting for it. And actually, you reminded me of, with the high school example, ProPublica put out an investigation a couple of years ago about firms that have popped up explicitly selling papers to people applying for undergrad, people that are trying to get into college, right?

17:22Reese Richardson: And they'll publish both in regular journals, as well as journals like the Journal of Youth Studies, or what have you, that it's entirely designed to never be read, but look like a peer-reviewed paper. There's no end to it, right?

17:38Rohin Francis: So something that I heard recently in cosmetic science is just to highlight that this isn't just like a others issue. It's also internally to us. What was surprising to me, and I don't know how this varies from PhD to PhD, but in their PhD, in the PhD cosmetic science program, should I say, there's a requirement to publish, but there's no requirement for the quality of journal. and so I've been talking to a number of cosmetic science PhD students like they have strong feelings for how quick peer review should be it's over a month way too long and here I am with my first thesis article that's been sitting in peer review for like half a year my publication has to be in a top tier journal otherwise I'm not allowed to publish finish your PhD and no publication sure that's fine.

18:32Rohin Francis: My advisor doesn't want me to embarrass her. But that's not the stance of all PhD programs. So their point of view, the PhD students are being like held hostage by the peer reviewers for the speed. Is this common? Do a lot of PhD programs have this requirement for publishing? And what is the possible problem for that? So it's obviously happening in medical programs, but where else is it happening? And what's the problem with that?

19:03Reese Richardson: It is quite common. So my PhD program, for instance, had a requirement for publication, I believe it was just one publication on which you were a first author, which is actually quite generous, right? Other PhD programs at Northwestern had similar requirements that were a lot more stringent about how many publications you had to put out. As a result, whereas in the past, it was not really expected that you actually publish something by the time you finish your PhD. It is now more or less guaranteed that you should do it in order to get your PhD. So a lot of places will not place a requirement on where you publish your work, which means that many students will get roped into publishing in a backwater outlet or somewhere where either no one's going to read it, or it's not actually going to be generally peer reviewed, or they're going to be charged an exorbitant publication fee for something that has no prestige or no quality associated with the peer review process.

20:07Reese Richardson: Other times, they require that the publication be indexed in something like PubMed, or Web of Science, or Scopus. These are selective indices. So there are nominally some forms of quality control. But the fact that so many places require publications to be in these indices also makes these indices a target for low quality publications. So a lot of paper mills, for instance, will specifically advertise that that will get you published in Scopus, Web of Science, PubMed. Moreover, being in these indices is not a guarantee of quality either. I could point you to many journals that are indexed in each of these indices that are garbage.

20:56Reese Richardson: PubMed in particular has sort of relaxed its rules as of recent for what it considers, and they may or may not have this disclaimer up anymore. The PubMed website has undergone a lot of changes. But in the last couple of years, they put up a disclaimer on every article saying that we're just a library. We have no guarantee over the contents of what's showing up here. The US government doesn't endorse it. You'll see a lot of anti-vaxxers talking about, we have this new publication out and it's indexed in PubMed. See, even the US government agrees with us. But yeah, the landscape for the requirements is quite diverse.

21:34Reese Richardson: In some places, it's quite stringent. In other places, it's not. I think it can be problematic to have requirements for publishing as well as having quite loose requirements for publishing either way.

21:45Rohin Francis: I'll give you a slightly different angle I didn't have a specific requirement for mine you know it was nice to get some publications but actually just doing the PhD in general I'll be entirely honest I probably wouldn't have done it had I not been in a field where it was almost expected of me so in competitive fields in medicine dermatology being a classic example but you know interventional cardiology, which is what I do, plastic surgery, neurosurgery, you know, these fields where you'll have often dozens of people applying for a single spot. To be competitive, once upon a time, you know, you needed, say, a bachelor's or a master's or something in addition to your MBBS, equivalent of MD here.

22:31Rohin Francis: But now PhD has become pretty much the norm. This has happened within recent memory from something that was a choice that people would do out of passion and interest to now feeling like, well, I've got to tick this box, or I've got to tread water for a few years while I improve my CV for a job. What kind of quality do you think is going to come out of the majority of people like that who are doing it as a means to an end? Having said that, I want to take a bit of a step back. And I think maybe I've been a little mean on the individual level that people are knowingly producing bad science. I don't think that's the case all the time.

23:06Rohin Francis: People are doing these kinds of PhDs where they've come out of their medical training, we all know something like a clinical trial, it can take years to set it up, go through ethics and recruits. It's very tough to do something really meaningful in a short space of time. You end up doing these proof of concept pilot studies, which can still be really beneficial to the person doing them, but I don't think they really push science forward in a big way. And so the problem lies not with the individual, but with the system they find themselves in. So I just thought I'd add that because I don't want to make it sound like it's all just disingenuous people doing it.

23:42Rohin Francis: But just the system, it would be great. People quote the sort of Fred Sanger argument, two Nobel Prizes, but he basically published nothing through the whole of his career. Just he won one early in his career and one at the end. And in between, he published almost nothing. So that era is gone where you can just kind of come into work and just do whatever takes your fancy for 30 years on end. But it would be nice to have that kind of curiosity and scientific pursuit of whatever you find interesting, useful without having that publish or perish. We haven't said the phrase publish or perish so far.

24:18Rohin Francis: Which is impressive. Yeah, but that's what turned me off academia because I saw my supervisors spending so much time worrying about grant applications and maintaining their job security by producing those five papers a year or whatever, instead of really sinking into what took them into the field, which is doing research. And it just seems like this perverse kind of conveyor belt where as soon as you gain some seniority, you actually get further away from the science that really brought you in. if they didn't have that pressure hanging over their heads the whole time, they'd probably do some much more interesting research where they don't have to produce papers on a yearly cycle, but maybe every five years, then something like that.

25:01Reese Richardson: So that's actually how the UK evaluation system works. I believe it's called the research excellence framework is they have these five-year deadlines. And there's some evidence that paper quality actually decreases a year before the deadline, right? Well, but you raise an interesting point, right? These metrics are compiled on a year-by-year basis as if science is something that happens on a year-by-year basis. And no, it takes decades. If we're talking specifically in biomedical science, it can take multiple decades for something to go from a basic scientific discovery to, say, an approved drug for a common disease.

25:44Reese Richardson: That's being generous to say that that process isn't a century or more in the making. Being able to demonstrate value in the short term has become very essential, but it's not really how the process works. And as a result of that, we've seen a reliance on simplifying heuristics, like how many papers you've put out, what's the impact factor of the journal you put about? What's your age index? How many patents do you have? Do you have a PhD in addition to your MD? Right? And all these things are very easy to measure, but are really only peripherally related to the things that we value in scientists and in physicians.

26:24Rohin Francis: Yeah, totally agree. And I guess, just for a quick rundown of the systems and science that are creating some of the issues. I'm just going to quick summarize. I had several podcasts on this topic already. Reese was part of one of them. I also had a conversation with Dr. Miu Yang, another research sleuth, and Dr. James Heathers. Definitely, if you want a lot more on what I'm about to say, go listen to those conversations. Also, Rowan and Reese, if you want to add to anything that I say, by all means. But in terms of publisher parish, researchers have to publish often to maintain their jobs, to maintain their funding.

27:10Rohin Francis: They have to have impact. But what's impact? That can mean a lot of different things, but maybe like get a lot of visibility in media outlets. Okay, well, how do I do that? Well, maybe I'll sensationalize the findings. Maybe not them. It's the university like press office. Now researchers are also doing this kind of thing on social media. And then that drives sensation. And I mean, the attention economy kind of factors into this. Same kind of issues we see driving misinformation on social media also has a play in like every other area, such as in science the 30 profit margin of some of these that's actually the thing that when i found that out made me so angry of the journal businesses the only people that get paid are not the scientists it's like the journals and the business people on top and so the peer reviewers are these poor scientists that care.

28:22Rohin Francis: They're doing it for free. I feel like people are advocating. I think they are. James, oh my gosh, I'm sorry. I'm pretty sure he said, I forget that conversation. It was a little while ago. But generally, peer reviewers aren't paid, and it's a lot of work. And scientists not only aren't paid, they pay if they want people to see their research. And then their research is taxpayer-funded often, depending on what mine is. and then the if the public wants to read the research and i haven't paid a lot to make it open access then they pay and the payment amount per journal is like thousands to like

29:04Reese Richardson: i don't know i think nature publishing thousands yeah we can check we can check the the number in real time so the apc for nature family journals or for nature itself for gold open i'm going to

29:18Rohin Francis: Let's guess, Jen. Let's guess. I'm going to say$13 ,000. Oh, we're close. We're close to each other. We're doing the price of the grade. Higher, lower.

29:30Reese Richardson: It is indeed$12 ,850. Oh, my God. You're incredibly on the money. So, yeah, I can talk about, like, the publishing industry, right? So, yes, they make an insane amount of money. And this started up in the 80s with a guy, Robert Maxwell, businessman that sort of jumped in and started to reshape the publishing industry.

29:58Rohin Francis: Also Ghislaine Maxwell's dad.

30:01Reese Richardson: Yeah, Ghislaine Maxwell's dad, which is an interesting footnote here. But especially with the advent of open access, where instead of universities buying subscriptions to academic journals and portfolios and authors instead paying to make these articles freely available, all of a sudden the way to profit has shifted from maintaining a valuable product to publishing as much as possible. Effectively, publishers make money in proportion to the number of articles that they actually publish. They do not make money in proportion to the number of articles that they reject. They don't make money in proportion to the number of articles that they retract.

30:45Reese Richardson: Yeah, basically, there's a incentive against doing better curation of the information ecosystem. And in fact, although publishers are investing more in having publication integrity teams to survey their published catalogs, these often pale in comparison to the number of professional editors and editorial assistants that they are employing to push papers out as fast as possible. And like you said, a large part of the actual value that they're adding or claiming to add is by facilitating the peer review process. But peer reviewers are usually not paid. So they're not compensated. They're expected to do it as part of their careers.

31:31Reese Richardson: At the same time, universities and research institutions typically do not value peer review as part of your portfolio when you're applying for tenure or applying for hiring or promotion or what have you. So yeah, there is a complete misalignment of incentives on the publishing side. And I'll also add that anybody that's published a scientific paper knows how annoying it is to actually interface with the publisher. You think that they're going to do anything and they're like, actually, your article's not formatted the way that we wanted it to be. There should be headings and subheadings and you've only provided headings or your citations are in this format and we want them to be in this format.

32:15Reese Richardson: OK, well, what are we paying you for other than to format this to your journal specifications and to host the PDF online? It's an incredible abdication of responsibility in most cases.

32:28Rohin Francis: I totally agree with all of that. And exactly this, that, you know, what are we paying for? That's always the, because if you've published for other kinds of publications, like, you know, regular newspaper or magazine, then the sub editor will do all that, you know, they will sort out your formatting and your style and, you know, often change the language and things like that. Now, in scientific publications, that's generally not the norm. But for things like formatting, and that is what staff on a journal should be doing. And as you mentioned, these are huge profit margins. The big five are vastly wealthy companies, and they've grown by buying up lots of smaller journals and then either cannibalizing or just ruining them.

33:14Rohin Francis: You're paying for some kind of quality control. but then if you just look at protraction watch and all the journal articles that have been taken back the the most uh cited ones and obviously the the biggest journals are going to have the most citations so obviously there's a bit of a skew there but these are in totally mainstream journals like obviously the most famous example being the lancet and the mr autism study now why are we paying so much why are readers paying so much to access these journals and libraries you know, a library license that all sort of different libraries have to pay. And a lot of libraries have said they simply can't afford to be subscribed to all the journalists anymore.

33:54Rohin Francis: What are they paying for? And, you know, the fees that authors pay, if it isn't quality control, if it isn't a sub-editor fixing all the formatting, I don't get it. What's this? As you say, it's just hosting on a website. They're not even printing paper anymore. Like, okay, maybe it's printing cards. It's just, you know, we can do it. So some people have proposed like moving to the preprint model that exists in physics and mathematics and coding. But of course, the big difference there is, you know, these publications are not likely to lead to physical harm in those other fields. Whereas, as you mentioned, anti-vaxxers and stuff like that, publishing some horseshit on a preprint server.

34:37Rohin Francis: And then a journalist or some blogger comes along and says, oh, you know, ivermectin can cure cancer and cure COVID. That's how you end up in that scenario. So I don't think that's a step necessarily in the right direction. But maybe it's a sideways step if the journals have become so bad that even the really kind of mainstream journals mess up. So I'm not sure what we're paying for.

34:59Reese Richardson: Yeah. And, you know, I'm pardoned by the emergence of what's called the Diamond Open Access Movement. And the idea behind this is instead of universities paying to publish research at point of service, a university library will fund the publisher such that anybody can read it. Right. So they basically buy into this this collective ahead of time, not just so that their authors affiliated with their university can publish in there, which is often not a requirement, but that other people can read it freely without having to pay. so this again like the publisher does not make money in relation to the number of articles that they're selling their set they make money in relation to their reputation right their reputation for curating good articles that people want to read there's one initiative that just launched recently that i'm very excited about called the open journals collective and this is a collective of non-profits, publishers, university presses that have organized their journals into this collective so that they can all be subscribed to at once, right?

36:11Reese Richardson: And it's mostly in the humanities, in the arts and humanities and social sciences right now, but they have recently started adding a lot more journals. There are up to 279 journals that I'm looking at right now that are in science and medicine. I'm heartened by that. But like Rohan said, it's a big question of what role do journals have in quality control. And it's clear that there are many conspicuous lapses in quality control happening even at what we consider the best, highest bar journals.

36:48Rohin Francis: And speaking of, This brings me to one of Michelle's thoughts. So she wanted me to bring up, you probably have seen some of the drama around benzene churning up in cosmetic products. So a lot of the headlines come back to publications. Now there was a whole bunch of issues with the various ways that they were publishing. Some of it was being pushed to the media prior to publication, but being pushed in journals. They had a lot of affiliates within their lab that were at universities like Yale. But the studies were published into the Journal of Investigative Dermatology. There were credible people peer reviewing, except they were dermatologists.

37:45And the topic of benzene churning up in cosmetics is analytical chemistry.

37:53Rohin Francis: topic that like how would they possibly how would they possibly do good peer review and then afterwards so I guess like now we're starting to get into like how is this impacting stuff the dermatology times editor-in-chief who collaborated with Fallisier presented these findings to a room full of dermatologists at American Academy of Dermatology conference And then from there, because nobody has the relevant expertise to vet this stuff, especially like, I don't know, you'd probably expect being in this room that the speaker is giving them credible information. They went in like there were dermatologists outside of all the crazy headlines that popped up from there.

38:43Rohin Francis: There were dermatologists firstly on social media making warnings to their followers about benzene and cosmetics, possible carcinogen. They don't even understand the toxicology behind the like exposure scenarios likely to cause an impact because like not their background. But then there are dermatologists going in calling their patients to warn them about the benzoyl peroxide products that they had recommended. And like benzoyl peroxide for like, I don't know, that was really important to me growing up as a kid, a teenager with acne. And now you're like, you're scaring your patients. And I mean, that's not something that the patients are probably going to quickly forget or the world because of factors like, like, continued influence, even if it were to be corrected by a relevant expert.

39:38Rohin Francis: I feel like some of the noise due to misrepresentation, which I think was honestly done, it's so sticky. For my audience, if you want a whole bunch more on just the valisher benzene topic, I had a roundtable on my last podcast. Michelle was on it, as well as a toxicologist. So Lyle Bergoon. So the two relevant expertises for this topic, just so people are aware. But what do you think about that story and like other possible kind of issues and outcomes? All I can add to that really is as a non-expert, but maybe that's the kind of theme of what I'm going to say is I think we're all familiar with the kind of Dunning-Kruger and people not understanding the limitations of their own knowledge.

40:26And I think that can be quite pronounced within medicine.

40:31Rohin Francis: I mean, I don't know if it's uniquely special medicine, but obviously that's the field I know best. You know, this understanding of different areas of expertise and maybe not even being aware that, you know, actually this is something that I would defer to a chemist to give a more informed opinion. And kind of coming back to what you were saying about peer review and just to explain to the audience, you know, the reason peer review exists is exactly for this reason that a journal staff, when you're looking at something like medicine, cannot be experts in everything. Even within cardiology, you have very distinct areas of expertise.

41:08Rohin Francis: So I wouldn't feel comfortable peer reviewing a paper on, you know, some kind of electrical heart problem that's very cutting edge when I'm a plumber. So you peer review is you're supposed to ask people with the relevant expertise, but that system can often be flawed. If the people asking may not know that people being asked might feel, oh, yeah, I can I can do this. But actually, they're not the right professional. or the not-for-right expert to talk to?

41:36Reese Richardson: I don't know much about the benzene situation, but it is incredibly common to see these panics sort of balloon out of control if somebody suggests that there is an association between A and B. And as we touched on earlier, these studies are becoming much more frequent. It's a classic to say like, oh, red wine does this thing, or red wine does that thing, or you should eat this many chocolates a day, or eating this many chocolates a day makes you more likely to have this, right? One recent snafu that comes to mind is the black spatula issue. So there was an article that came out that was basically saying that flame retardants that appeared in black cookware, black plastic cookware, was leeching off into food and was reaching incredibly unsafe levels.

42:27Reese Richardson: This went crazy in media and basically led to a bunch of people saying, like, I'm throwing out all my black plastic cookware, right? And I'm buying this brand new thing, which very quickly, probably as a result of the immense attention that this was getting, people started to look into the original study and found, wait a second, they made several errors that led to them overestimating the magnitude of leaching and the magnitude of human exposure of this particular chemical. And the study wound up being corrected several times. I don't think it actually ever wound up being retracted, but it got so significantly corrected that everything that it originally claimed was inaccurate.

43:11Reese Richardson: And sort of interestingly and uniquely for the study, a bunch of outlets then went to cover the fact that, you know, we were wrong, which I think was quite rare, something that doesn't really happen with these sort of things. Even when something is retracted, a lot of the time people will not come back to cover it in that light. Obviously, we brought up the Wakefield paper earlier. One thing that's interesting about that particular case is that Wakefield actually had patent applications out for separate MMR vaccines. His big push was these can't be combined. It's too much at the same time. And that's ultimately what all the press conferences and all the media coverage was about, was we need to split these things up.

43:59Reese Richardson: So he stood to gain financially from these things. And this is something that we see quite frequently. So with the, you know, benzene research, one thing that you might want to be looking out for, not saying that this is actually happening, I don't know enough about the situation, but certainly there is no shortage of these sort of things occurring, is the research that is promoting the narrative that there's particular harm in this being funded by somebody that's trying to sell an alternative? Or, you know, it could be the other way, you know, is research that is showing that there is no particular harm to exposure to a toxic chemical being funded by a manufacturer of that chemical.

44:45Reese Richardson: There is not necessarily one media spin and sort of media frenzies can lead to a lot of panic, but a lot of this panic can be specifically manufactured to promote a particular narrative.

45:00Rohin Francis: Now, this may be a controversial question, but I think it's important. And it's been something that's been coming up a lot in the past few years. It's hard to praise the literature, whoever you are. so even if you're a PhD scientist or highly trained scientists doing research as soon as you're in a space that's not your own this is like a reoccurring theme on my podcast you're going to have a really difficult time in that because you have to understand the specific methodologies you have to be able to put the study into context of the whole body of peer-reviewed literature see if it actually makes sense has it been replicated and so on and so forth there's a lot that goes into critically appraising a study.

45:48Rohin Francis: And it's hard even for people who are in that space. It's almost, it's not impossible, but very, very, very hard for somebody who's not steeped in that literature. Okay. So like groundwork is hard for everyone, but then, so there's a lot of dermatologists that seem to think like if it's peer reviewed and published, then therefore it's good. And like, as an example, there was a whole bunch of dermatologists talking about how rosemary oil is going to fix people's hair loss. And they were saying that loudly on social media. Like I would like people to know that. But then when corrected, they would push back and say, especially when Michelle and I had questions about like, did you read the study?

46:32Rohin Francis: Do you understand how it works? And so on and so forth. Stop being a hater, Jen. Do they know how to do research. A commentary that I've said on, I hosted a podcast to talk about some of the stuff that I've seen in the dermfluencer space, dermatologist influencers. That's what dermfluencer means. Skinfluencer was right there, Jen. You've gone for dermfluencer. Well, it's specifically the dermatologist. Okay. All right. So I hosted a round table with, it was from the point of view of the lawyer, as well as a dermatologist who does research, Dr. Ade Adamson, which was what the fuck is up with dermfluencers?

47:13Rohin Francis: I posted a carousel recap on my Instagram. And then I highlighted the fact that PhD training is different than MD training. They're not trained to the same level. This is my point of view and the conclusion of that conversation. There's so many people in the comments who were dermatologists, dermfluencers actually. I think like the non-dermfluencer dermatologists in the comments where a lot were in favor of the content of the podcast, which maybe says something, but they're like, no, absolutely wrong. Dermatologists do get a lot of training in how to do research. Okay. So like the question is like, do they?

47:50Rohin Francis: And Rowan, you're uniquely positioned here since you've done a MD and you've done a PhD. So like, what's the difference? How robust is the training of MD for critical appraisal? Am I incorrect to say, because they talk about, we have reading clubs for the studies to appraise them all together. As far as I understand, they're all just basing it off of the hierarchy of evidence. And that's the way. If it's a certain type of study, it's good. If it's another type of study, it's bad. And like, that's it. or not bad, lower quality. And that's it. But there's so much other stuff going on there for critically appraising.

48:34Rohin Francis: Okay, I'm going to stop talking. Rowan, what do you think? I'll start just by referencing a funny thing. Do you know Darren Daly? He's a biomedical statistician. So he and one of his colleagues, whose name I forget, I apologize, came out with a new pyramid of evidence. So instead of the famous, you know, meta-analysis, systematic analysis, and all that stuff. They just had a pyramid where the top is good studies, and the bottom is all the other shit. And the point is that you can have a really well-designed cohort study or an observational study, which gives you some useful information. It doesn't necessarily answer the question that the media might want to know, but you can still get a lot of good information from a well-designed study that isn't a randomized controlled trial.

49:21Rohin Francis: But for certain questions, you do need a randomized control trial. But systematic system, system, analysis of other papers, I'm having problems speaking, meta-analyses, you know, garbage in, garbage out. So if you are amalgamating a lot of bad studies, it's going to be. So I think in answer to your question, some people will use that as a crutch. They will just look at that and go, it's a meta-analysis, it's a systematic analysis. And that's the be all and end all. How much are doctors taught? Well, I can't speak for every medical school curriculum, of course, but I did get some training at medical school.

49:57Rohin Francis: We did have journal clubs. I think some of that has been eroded in recent years. I think the pandemic really did a number on those kind of in-person meetings, which were so valuable. We'd have weekly meetings in the hospital where we'd discuss papers and stuff like that. That's kind of died off, really, I'm afraid to say, in almost every hospital I know of in the UK, at least, where you could meet and chat about studies that are of interest. and the people who really like doing that now I think are a bit of a niche area and where you may learn the basics it's a it's a muscle if you're not flexing it on a regular basis you're it's going to atrophy and then you forget I'm going through a paper is it does take work you know it does take a long time to really go through something as you said if it's slightly outside your wheelhouse you know I can read a paper in neurology and you know I'll spot big glaring problems and maybe some obvious statistical issues, but I'm no stats guru.

50:53Rohin Francis: But subtle things like, you know, the typical clinical practice for what is done in a neurology clinic, I'm not going to know that. So to really analyze a paper takes a long time. And it's not like some kind of savant that you can just sit down and understand any scientific paper, no matter what the kind of Elon Musk's and people like that think that you can just apply first principles. You know, I think there are some responsible people. I don't think we can generalize too much. I think there are plenty of doctors who simply think, well, hey, listen, I'm not the expert in analyzing data, but I've got professional bodies that produce guidelines.

51:30Rohin Francis: And I think the majority of doctors who say, I'm going to look at my society guidelines, I'm going to look at the National Institute or the Royal College and see what their expert consensus opinion of my peers is. And I think that's a very straightforward, sensible attitude to take for most doctors because they're busy seeing patients. They're not sitting there going through reams of data. And the few who are really motivated to go into academia, they build those skills. And I feel I'm still learning. And I still think I'm pretty mediocre when it comes to reading a paper. But I think I'm a lot better than the average kind of influencer.

52:08Rohin Francis: and I think the ones that are most high profile are often the ones with the most rudimentary knowledge. Yeah, it is a major problem. I know from just to come back to what you were saying about you and Michelle's work, I'm so grateful for it and the rosemary oil one in particular, Michelle's video on that, I thought was really one of her best. It was just such a great tour of a lot of the problems and the mistakes that people make. But why I value what you do so much is because you take on those harder targets. And I think a lot of science communication these days is just floating heads, doing sassy clapbacks to things that are, there's a role for that.

52:48Rohin Francis: It can be useful. And it's very entertaining, of course, to just make fun of some influencer talking rubbish about their carnivore diet or screaming in a supermarket with no shirt on. I think the majority of people already don't believe that thing. It can still be beneficial to understand why it's not true for sure. But taking on the harder targets is so much rarer because as you say, it takes work, it takes knowledge, it takes precinct knowledge. You know, I'd like to think I try and do similar things and, you know, try and dismantle a lot of the very credentialed kind of influences. But yeah, there's a lot of pushback because I think they're expert communicators.

53:27Rohin Francis: They're very good at getting their message across and they sell it in a way, you know, I'm talking to two friends at the moment who deal with a lot of menopause misinformation. And, you know, when they've written articles about this in the British Medical Journal and stuff, they get attacked. I mean, they're two women, but they get called misogynists and things like that because they are denying women from these life-saving treatments. And, you know, they don't want menopause to be taken seriously. So there's a lot of emotion tied up in these things as well. Rhys, do you have anything to add?

53:57Reese Richardson: I in particular, like you mentioned the evidence pyramid, right? And you alluded to the Rothman Daily evidence pyramid. Rothman, yeah, sorry. Yeah. So basically, this is the idea that like a systematic review and a meta analysis have primacy. Like most things, a systematic review and meta analysis are subject to garbage in garbage out problems. So for instance, there was a Cochrane meta analysis, which is sort of the gold standard in meta-analysis is done under the auspices of the Cochrane collaboration that was about use of a particular drug for post-menopausal women. I can't quite remember the details, but now it's becoming more and more frequent that Cochrane meta-analyses are changing their outcomes, the outcomes and the recommendations of the meta-analysis based upon, these studies have been retracted or these studies are not reliable.

54:54Reese Richardson: Some colleagues of mine created a tool called Inspect SR that is about assessing the trustworthiness of studies that go into a meta-analysis. And it's beginning to be more and more used. But I'll also note, meta-analyses are a really good study to do if you do not have time and you do not have resources or you do not have your own lab space. In fact, many of the studies that were being sold by these paper mills catering specifically to international medical graduates were meta-analyses because they required no new collection of data and they were relatively easy to do and pass off. And many of them were quite poor quality.

55:38Reese Richardson: I think of one in particular that vastly overstated the harms of vaping in relation to smoking that was recently retracted. And it had like 20 international medical graduates on it. The lead author was the person that was running this business, selling publication opportunities to these international medical graduates. So yeah, it's a matter of, I don't want to say like, do your own research, right? Because that is a problem in itself, but it's a complex information landscape. You should not be trusting any one single expert based upon their opinion. That's sort of the point of doing high quality peer review is running it past not only the original authors and the editor, but also a slew of people that can look at a piece of evidence from multiple angles and assess its credibility and trustworthiness.

56:33Rohin Francis: Do you have any final take comes that you'd like to leave my audience with? I'd say unfollow 95 % of health influences and follow more accounts about nature and gardening and birds and interesting things like that. Biology, evolution, much more interesting than handling receipts or sunscreen ending up in your brain? I was thinking about advice that I could offer to anyone just in general.

57:07Reese Richardson: And it's sort of to know that the truth is very complicated. To leave room for nuance, to understand if you were, you know, encountering a specific narrative that is not very well justified and you're not seeing a lot of evidence for, you may be undersold. People should know that when they're encountering something on social media, they are seeing a very simplified version, something that has been sort of packaged to be simple and to be easily digestible that doesn't necessarily represent the whole truth.

57:46Rohin Francis: Thanks again to Rowan and Reese for joining me for this episode. I hope you enjoyed the discussion. quick reminder if you're not already subscribed to the podcast please do and if you haven't already please give the podcast a review if you enjoyed this episode please share and if it's accessible you can also support the podcast on patreon at patreon.com slash the eco o thank you again to all my patrons supporters who supported my sci-com efforts over the years that's all for this week thanks for tuning in