Undisclosed Financial Conflicts of Interest in the DSM-5: An Interview with Lisa Cosgrove and Brian Piper

On the podcast this week we turn our attention to conflicts of interest (COIs) and new research from the British Medical Journal (BMJ). Mad in America has previously examined the problems with conflicts of interest in research but this time we extend that to look at the potential effect of COIs on diagnostic tools such as the Diagnostic and Statistical Manual of Mental Disorders (DSM).

Joining me today are Lisa Cosgrove and Brian Piper, two of the authors of a paper which appeared in the BMJ. The paper is entitled “Undisclosed Financial Conflicts of Interest in the DSM-5 TR: Cross-Sectional Analysis,” and it was published in January 2024.

Lisa Cosgrove is a clinical psychologist and Professor at the University of Massachusetts, Boston where she teaches courses on psychiatric diagnosis and psychopharmacology. A former Research Fellow at the Edmond J. Safra Center for Ethics, Harvard University, her research addresses the ethical and medical-legal issues that arise in organized psychiatry because of academic-industry relationships. She is co-author, with Robert Whitaker, of Psychiatry under the Influence: Institutional Corruption, Social Injury, and Prescriptions for Reform.

Brian Piper is an Associate Professor of Neuroscience at Geisinger Commonwealth School of Medicine in Scranton, PA, with a secondary appointment at the Center for Pharmacy Innovation & Outcomes with Geisinger in Danville, PA. He maintains an active program of research in the pharmacoepidemiology of controlled substances including opioids, cannabinoids, and other controlled substances, behavioral neurology methods development and quantitative medical ethics.

The transcript below has been edited for length and clarity. Listen to the audio of the interview here.

James Moore: Lisa and Brian, welcome both of you. Thank you so much for joining me for the Mad In America podcast. I’m delighted to have you on to get to talk a little bit about your work.

Lisa Cosgrove: Thank you. We’re delighted to be here.

Brian Piper: Thrilled to be here.

Moore: To get us underway, could tell us a little bit about yourselves and your areas of research interest?

Cosgrove: I’m a clinical psychologist by training. I’m also affiliated with the Ethics Center at the University of Massachusetts, Boston. The research that we do broadly looks at the ways in which academic-industry relationships can have a corrupting influence on medicine and, in particular, on psychiatry and psychiatric diagnostic and treatment guidelines.

Piper: I’m an Associate Professor of Neuroscience at Geisinger Commonwealth School of Medicine. I also have a secondary appointment with the Center for Pharmacy Innovation & Outcomes with Geisinger in Danville, Pennsylvania. I have two lines of research. One is on conflicts of interest in medicine broadly, looking at influential information sources that are used by physicians and other healthcare providers. Then I also have another line of research on pharmacoepidemiology, looking at the increasing use of medications, whether it’s for attention deficit hyperactivity disorder, depression, or for a wide variety of other psychological disorders.

Moore: We’re here today to talk about conflicts of interest, and in particular, a piece of work that you both did examining financial conflicts of interest and the DSM. But before we get to the paper itself, I think people probably have a broad sense that conflicts of interest can be bad. But I wonder if we could talk about specifically why they can be such a problem?

Piper: If we think about the history of psychiatry for a little bit, people are pretty well acquainted with Sigmund Freud and his very close relationship with cocaine. Cocaine is mostly associated with being a recreational drug that is smoked, but it also was and continues to be used as a medical drug, particularly for eye surgeries. Sigmund Freud was a major advocate for it. He wanted his friends to use this drug, he wanted his family members to use this drug. He was also a major user for an extended period.

When medical historians say, “This particular person got an appreciable amount of money from a pharmaceutical company, in this case, Merck and Parke-Davis,” we want to know how much money it is. When he says that there are no cravings associated with cocaine, we want to know, is that advice coming from his expert opinion having used this drug quite extensively, or do those conflicts of interest and the money that he was receiving influence the medical care that he was providing? That was an issue then, and that continues to be an issue to this day.

Cosgrove: The only thing I would add to Brian’s excellent point is that conflicts of interest can incur implicit bias. It’s part of the human condition to have biases and to remain blissfully unaware of them. We often assume, quite wrongly, that the scientific process can protect us from those implicit biases. So, one of the reasons why conflicts of interest in medicine are problematic is because they do lead to these implicit biases, what some have referred to as pro-industry habits of thought. In terms of the DSM, it hasn’t been that long that we’ve had this medicalized view of emotional distress, and the medicalized view of emotional distress is certainly an industry-friendly perspective because it leads to the development of psychotropic medications as interventions, but it’s not necessarily in the public’s best interest.

Moore: And of course, some of the people involved are what are known as key opinion leaders, aren’t they? So, they’re quite important figures in the field of psychiatry and one gets a sense of them acting almost as a marketing tool rather than a tool of science.

Cosgrove: Yes, that’s a very important point. There are key opinion leaders in all areas. Nike knows that to sell sneakers they’re going to tap a very famous athlete and in much the same way that Nike knows that, so too do the pharmaceutical company executives.

Moore: Let’s turn to your BMJ paper, “Undisclosed Financial Conflicts of Interest in the DSM-5 TR: Cross-Sectional Analysis.” What was the question that you were trying to answer with your work and how did you go about the research to answer that question?

Piper: I can start with a little broader perspective on the research question. So I’m coming at it from the pharmacology side of things. I teach pharmacology for medical students, for residents, and for graduate students, and influential materials can have a very profound effect on the practice of medicine.

Members of my team have been doing studies for the past 10 years, looking at point-of-care databases. We’ve been looking at a lot of different information resources, like the New England Journal of Medicine and the Journal of the American Medical Association, and the general pattern is that the conflicts of interest are really important. The National Academy of Medicine says, that if you’re going to put together clinical practice guidelines that you can trust, you should have authors that don’t have a conflict of interest wherever possible.

The American Psychiatric Association (APA) and their publications require that authors disclose conflicts of interest. So we were a little bit surprised, maybe even a little bit mystified that the APA, in publishing the most recent version of the DSM, didn’t have some sort of preface about the conflicts of interest, even in small font. That information wasn’t available. So, we wanted to help them out, if you will, and get this information into the public domain.

Moore: And it was the Open Payments Database you used, is that right?

Piper: Yes, it is an amazing resource that has been available for the past few years. At the risk of a digression, I would suggest to every listener: Before you go to your primary care provider, your psychiatrist, your physician assistant or your nurse practitioner, do a quick Google search for CMS Open Payments, and with a little bit of practice, ideally, if you know their middle initial, you can look up their conflicts of interest. And that’s what we did for the 160 or so contributors to the DSM.

Moore: If these people are aware that anyone can go and find that information, and they’re obviously happy to accept industry money, and for it to be presented as public information, do they not think it’s an issue? Do they not mind that this information is out there for researchers to grab and present to the public?

Cosgrove: That’s a good question. It is curious because when Shelly Krimsky and I first did our study in 2000 (published in 2006) on conflicts of interest in the DSM-IV, we found that almost 60% of the authors had financial conflicts of interest. Then, a year before the DSM-5 came out in 2012, we published a study looking at the conflicts of interest and, at that point, the APA did institute a conflict of interest policy. I’d like to think that maybe our 2006 study had a small role to play in that. At any rate, they did have this policy. And yet, what we found was approximately the same percentage of panel members had ties and the ties were particularly strong in those areas for which pharmacological treatment was the first-line intervention.

So as Brian was saying, it is curious that, given our past publications and their conflict of interest policy, they didn’t have one for the 2022 DSM-5 Text Revision (DSM-5-TR). What’s unique to this study, even though in many ways it’s a replication of our previous work, was that we were able to use the Open Payments database which started in 2014. Shelly and I weren’t able to do that in 2006 or 2012. So what Brian was saying is so important because as long as you’re in the US, you can look up any physician and see not only how much money they’ve received, but the type of funding and, as you were referring to James, you can see if they were, in fact, key opinion leaders.

Moore: Thank you so much. And can you remind me of the total amount that you found again? Was it $14 million?

Piper: It was $14 million, completely undisclosed. So the readers of the DSM wouldn’t know this. Unlike in the APA journals where there is this information, or at least you know the names of the companies, none of this information was available.

Moore: In reading the paper, one thing that struck me as really important was that if there was some regulation mechanism at the diagnostic end, that might help matters. If there were an objective biological test for example, that itself might regulate the increasing trend in diagnoses. But because we appreciate that psychiatric diagnoses in the main are quite subjective things, then if someone is looking to the DSM and doesn’t necessarily realize that there’s been industry influence, there’s nothing at the back end to help regulate the number of diagnoses coming through. Is that right?

Piper: Yes, that’s completely right. So, depending on how you count it, of the 300 or so disorders in the psychiatric bible, the DSM, there’s only a handful that have an objective laboratory measure. That includes defining an objective laboratory measure pretty broadly. You could include something like a scale like a weight scale for anorexia or, for narcolepsy, hyper creatinine deficiencies occur in about a third of patients. But for the vast preponderance, this is based on very expertly, very carefully obtained but subjective information.

Cosgrove: There are actually no biomarkers for any of the DSM disorders. So Brian is right and I just want to emphasize that point.

Moore: In the paper, you’ve done a breakdown of how much money was coming in per group of disorders. It was quite interesting to see that the biggest one was medication-induced movement disorders. Which I guess speaks that there seems to be an avenue there that the pharmaceutical industry is quite interested in exploring, and therefore is trying to maybe pump money into it to stimulate some demand.

Cosgrove: Yes. And again, there are two points I’d like to emphasize. One is that, once again, what we found was that the categories for which there was the most industry money were the categories of disorders for which pharmacological treatments are the first-line intervention.

But the other point that I want to emphasize is that we’re not in any way suggesting wrongdoing or any sort of quid pro quo mechanism at play here. Again, the issue is implicit bias that creates this pro-industry habit of thought which reinforces a medicalized view of emotional distress. So it’s not the case that we think that there are any sort of backdoor deals but, rather, by medicalizing distress, we deflect attention away from the upstream causes such as austerity policies, unsafe or inadequate housing, food scarcity and so on.

Moore: Is this a problem that’s particularly prevalent in psychiatry or does it affect all sectors of medicine?

Piper: I would argue it’s a really broad problem that isn’t limited to psychiatry. We’ve done other work looking at cardiology and oncology journals. This is a broad issue. I continue to do these studies beating on this drum that self-reporting conflicts of interest doesn’t work. I submit these manuscripts and some of the anonymous reviewers come back and say “Yeah, we’ve known for a decade that self-reporting doesn’t work and that mandated reporting is the way to go.” However, journals continue to rely on self-reporting. And I don’t want to disparage my gender at all. However, people with a Y chromosome have a particular difficulty with this issue.

We keep on seeing this over and over in studies. We’ve looked at contributors to the point of care database like Medscape. So we’ll look at it up and it’s like, oh, you’ve received money for the major products for Parkinson’s disease. You write an article on Parkinson’s disease and even though you receive $400,000, you say you have no conflicts of interest. That’s curious.

This is a widespread issue and I’d argue this isn’t just an American issue. We really should be grateful that America has a tool like CMS Open Payments that’s been mandatory for a decent period. In some other countries, it’s either more recent or non-mandatory.
So in some ways in America, we have our challenges, but at least we have some tools to begin to deal with this whereas some other countries are less further along in having tools like this.

Moore: Understood. Your study appeared in the British Medical Journal, which is a pretty prestigious publication. It only came out in January, it’s March now, but I wondered what kind of feedback you’ve had since you published your article.

Cosgrove: The APA did write an editorial immediately after it was published and they didn’t disagree with any of our findings. They tried to say that these conflicts of interest didn’t affect them, which, again, is problematic, because it’s just part of the human condition to have implicit bias.

The other thing I just wanted to circle back to thinking about what Brian had just said is what’s unfortunate are the downstream effects of the conflicts of interest. By that, I mean the fact that psychiatrists and other prescribers are not receiving balanced, accurate information. We know that there’s certainly a connection between industry funding and pro-industry results. And unfortunately, the peer review process is not robust enough to protect against that.

So like Brian, I teach a psychopharmacology class to masters and doctoral students. We did a study a few years back looking at conflicts of interest in psychopharmacology texts, and they are rampant. So one has to be concerned about our medical students and doctoral students receiving accurate and balanced information from the text. I would argue that when there are lots of conflicts of interest on the part of the authors, we need to be skeptical about the information that’s being disseminated.

Piper: Lisa’s done a fantastic series of studies on this topic. And we did get some interest from media organizations and podcasts like this. I had some particularly interesting comments on Reddit which is a fairly snarky place. Some of the people were just like, “Ho-hum. We know this, kind of yawn”. So yes it is getting out to the general public but this is an unfortunate situation where if the APA loses trust, it’s really difficult to gain that back. There’s an easy solution to prevent that, there are tools like open payments if they can list the link in the next version of the DSM. I’m young enough and Lisa is energetic enough so we will be there to do this for the next version of the DSM but we hope we don’t have to, we hope they will do that for us.

Cosgrove: To Brian’s point about the sort of Ho-hum response. If you told people that eating a candy bar increased longevity and overall health then they found out that that study was funded by Hershey, most people would be certainly skeptical. They wouldn’t have to be schooled in conflicts of interest. And yet, sadly, when it comes to medicine, I don’t think people are as aware as they could be of the ways in which these conflicts can result in imbalanced and sometimes inaccurate information.

Moore: You talked there about the next version of the DSM. So if we look to the future, is it sufficient to have these conflicts writ large in the work? Or should we go further than that, and examine some split between industry and how it contributes to research? Is it enough just to name these things or should we go further?

Cosgrove: Definitely. We should go further.

Piper: Just disclosing isn’t sufficient, it’s the management of these conflicts of interest that is important. The National Academy of Medicines’ dream would be to use people who don’t have conflicts of interest. Also, as the former chair of the DSM-IV task force has indicated, 80% of psychiatric meds are prescribed by general practitioners. So why not have general practitioners and other people that are involved in this process? So no conflicts of interest and other voices on those DSM working groups and panels to have those other folks better represented. That would be my dream.

Moore: Is there any reasonable chance of that happening?

Piper: I’ve been a little underwhelmed. I’m a neuroscientist who came from the decade of the brain and was excited to be learning more about the biological underpinnings of psychiatric disorders. And even when there’s been progress in some psychiatric disorders and genetic components, the number of geneticists on the DSM panels is extremely limited, it continues to be underwhelming. So I’m not optimistic, frankly.

Cosgrove: I think the DSM could be a more helpful diagnostic instrument if you included the people who are affected by the conditions. We have to remember that the creators of the DSM don’t just have the authority to change symptomatology or add a new disorder, they also have the authority to narrow diagnostic categories or even take out previously included disorders. Homosexuality would be a good example.

Moore: Was there anything else we should talk about that I hadn’t got to?

Cosgrove: Two things come to mind. One is the issue of guild interests. As I was mentioning, that’s why I think if you have a diversity of people on the panels, and on the task force, you’re going to get a more well-rounded instrument. What’s that saying… Never ask a barber if you need a haircut, a psychologist if you need psychotherapy, or a psychiatrist if you need medication. So the guild interests are another inevitable part of what it means to be human. I wouldn’t trust a guideline on psychotherapy if it was written by all psychologists. You need methodologists and you need more than just content experts. You need people who don’t have skin in the game, in terms of what they’re recommending.

Then in terms of the explosion of diagnoses, there was a really interesting article that came out a couple of months ago on what’s called the prevalence paradox. One of the things that the authors pointed out is that in areas of medicine other than mental health, when you have good treatments, then the prevalence of the disorders goes down. Antibiotics would be a paradigmatic example. Infectious diseases and bacterial infections went way down, a real game changer for humanity.

We don’t see that in the mental health field. We have more treatments, but we have an increasing number of people with disorders. Now I think the reason for that is multi-determined, it’s certainly not just related to industry and conflicts of interest. I think that social media has played a significant role, particularly in the increase in ADHD diagnoses. And there’s this part of the zeitgeist now, many people are encouraged to identify with disorders in a way that I don’t know is necessarily helpful for humanity.

Piper: I have just a big-picture thought. Lisa and I played an important role in this but there are others as well. I want to shout out for Lauren Davis and students at Geisinger Commonwealth School of Medicine and students at UMass Boston who played a key role in collecting and triple-checking this information and making this publication happen.

Moore: It’s clear from reading the paper how much effort went in and I’m so glad to see it in such a well-read place as the British Medical Journal. Thank you both for taking some time to explain your work and approach.

Cosgrove: Thank you so much.

Piper: Thank you.

***

MIA Reports are made possible by donations from MIA readers like you. To donate, visit: https://www.madinamerica.com/donate/