Top Selling Drugs, Conflict-of-Interest and Market Hype

 

This post is another public service announcement to the criticism psychiatric medications in perspective. It is an essential part of the continuous production of antipsychiatry rhetoric because it must be.  If you are arguing that there are no such thing as psychiatric disorders – you do not need any treatment for them. On that basis overprescribing easily occurs because any prescribing is overprescribing. There are mixed agendas on claims that antidepressants in general cause increased number of suicides and aggressive behavior. But the most significant rhetoric has always been that there are massive conflicts of interest in psychiatry due to payments or relationships with the pharmaceutical industry. In a rare piece of special interest politics, a US Senator made these claims. Now that we have a database, it is easy to show that psychiatrists in general are not even close to the top when it comes to physician payments from the pharmaceutical and medical devices industries and in fact that most physicians have no significant conflict of interest.  This post specifically is about the top 15 selling drugs in the period 1990-2021, and where drugs typically prescribed by psychiatrists end up on that list.

I will qualify this by saying as I have many times in the past, that the only medication that psychiatrists prescribe more of than any other medical specialty is lithium.  Practically all other classes of psychiatric drugs are prescribed in greater amounts by primary care physicians, pediatricians, and an ever-increasing number of non-physician prescribers primarily nurse practitioners and physicians assistants-certified.

The methodology for this post is subject to several limitations.  First it depends on word and term frequencies counts from search engines that look at the popular, press, academic journals, newspapers, websites, and books.  By comparison the Google NGRAM search looks only at frequencies in their book collection. Algorithms are used to determine relevance and importance – but the weighting of the algorithm is not available so the actual search is a black box – but consistently applied.  It does not count word frequencies in the non-print media and in the US that would include direct to consumer advertising for many of these drugs. A second limitation is that all the counts obtained here were for the years 1990-2021 in aggregate. No curves were generated by year like the ones available in NGRAM. Sometimes a verbal estimate was available like “most of these references occurred in the past 10 years”.  A third limitation occurred when searching for terms like “antidepressant overprescribing” the search engine went from Google Scholar to Lexis-Nexis even though I had no direct access to Lexi-Nexis.  I have a request pending for that access to see if I can confirm the results listed here, especially because much of this search required an artificial intelligence search engine (Google Bard).  If I cannot get direct access to Lexis-Nexis I will attempt to check the results with a second AI search engine.

My goals were to compare the general publicity with the negative publicity of the top 15 selling medications of all time.  When I looked at that list – only one of the 15 medications was a psychiatric medication and that was olanzapine (Zyprexa). For that reason, I decided to extend the comparison to all antidepressants that were discovered during the same years (1990-2021).  During that time there was continuous criticism of psychiatrists that was primarily antidepressant based. The suggestion was that all psychiatrists were in lock step with Key Opinion Leaders (KOLs) and prescribing whatever Big Pharma wanted them to irrespective of patient need or diagnosis. This is obviously an absurd criticism but it was taken very seriously for the better part of two decades. An associated criticism was that clinical trials of psychiatric medication were permanently tainted by the same conflict of interest.  Psychiatrists were mere puppet prescribers for the pharmaceutical industry.  Like most dichotomous arguments there were vehement supporters of both poles but very little discussion of the reality that psychiatrists are the treatment providers of last resort dealing with difficult problems and they just happened to have the most expertise in prescribing these medications. These arguments also ignored the fact that psychiatrists are taught to critique literature and apply it to practice and that this life-long skill makes it doubtful that anyone prescribes a medication just based on the word of a salesman. All the attacks on monolithic psychiatry ignored what was potentially the most significant cause of antidepressant prescribing – checklist-based diagnoses popularized by managed care and electronic health record systems. No comprehensive psychiatric evaluation – just a cutoff number on a checklist. And yet no criticism of managed care companies doing this?  I have also not heard the same level of criticism for telepsychiatry services that are currently advertising their antidepressant prescription services.

 

Table 1. Top 15 gross selling medications of all time:

Medication	Sales (Kiplinger 2017)	Sales (Bard 2023)	Word Frequency References 1990-2021(1)	Overprescribing by drug references1990-2021 (1)
Lipitor (atorvastatin) 2001	$150 B	$130B	38M	20,000
Humira (adalimumab) 2008	$109 B	$240B	17M	1,000
Advair (fluticasone/salmeterol) 2000	$95.7 B	$350B	12M	1,000
Remicade (infliximab) 1998	$90.3 B	$100B	2.5M	100
Plavix (clopidogrel) 1997	$83.7 B	$55B	13M	10,000
Rituxan (rituximab) 1997	$81.6 B	$60B	1.3M	100
Enbrel (etanercept) 1998	$81.2 B	$100B	5.7M	300
Herceptin (trastuzumab) 1998	$70.5 B	$140B	1.2M	100
Avastin (bevacizumab) 2004	$67.4 B	$100B	1.5M	100
Nexium (esomeprazole) 2001	$61.8 B	$150B	7M	20,000
Lantus (insulin glargine) 2000	$61.6 B	$100B	1M	1,000
Diovan (valsartan) 2005	$60.8 B	$100B	2M	200
Zyprexa (olanzapine) 1996	$60.6 B	$55B	3.5M	1,000 10,000 for APs
Crestor (rosuvastatin) 2003	$56.9 B	$125B	1.5M	20,000
Singulair (montelukast) 1998	$47.9 B	$50B	200K	100

 

Table 2. Antidepressants approved 1989-2021

Antidepressant	Total Sales	Literature references (1)	Overprescribing by class 1990-2021 (1)
Wellbutrin (bupropion) 1985	$40B	2.5M	100,000
Prozac (fluoxetine) 1987	$32B	22M
Zoloft (sertraline) 1991	$50B	16M
Paxil (paroxetine) 1992	$40B	13M
Effexor (venlafaxine) 1993	$70B	10M
Celexa (citalopram) 1998	$50B	12M
Lexapro (escitalopram) 2002	$40B	8M
Cymbalta (duloxetine) 2004	$35B	14M
Viibryd (vilazodone) 2011	$1.5B	3M
Trintellix (vortioxetine) 2013	$2B	1M

1.  Lexis-Nexis search via Bard.

 

Table 3.  Approximate Number of Prescriptions

Drug Class	Estimated Number of Prescriptions
Statins (atorvastatin, rosuvastatin)	200M worldwide
Antidepressants (table 2 plus earlier ADs)	150M worldwide
Monoclonal Antibodies	Millions
PPIs (esomeprazole)	17.9M in 2008
Clopidogrel	18M in 2016
Valsartan	18M in 2016
Advair	2.5M worldwide
Etanercept	2.5M worldwide

 

A couple of qualifiers before comments.  First there is no organized registry or pharmacosurveillance system for basic monitoring of any prescriptions in the United States and in most countries. Total prescriptions and sales in the US have always been proprietary data.  It typically involves sampling local pharmacies and then extrapolating those sales to the larger population. Practically none of that data is available to the public but it occasionally shows up in articles in both the popular and professional press. Second, the proxies chosen here for comparing drug classes including the number of times a particular drug or term is mentioned in the popular and professional literature is also approximate because these words can be used frequently in the same paragraph and the context is not apparent – the words could be used in both positive and critical contexts. The “overprescribing” term closely paired to the drug name was chosen as a proxy for critical articles – but even then, the level of criticism and whether it was appropriate or not is an open question.  Looking at the general number of people exposed to each drug class it is reasonable to assume that 5 of the drug classes – statins, antidepressants, PPIs, clopidogrel, and valsartan have probably all been prescribed to about 200M people. The immunological drugs including the monoclonal antibodies and etanercept have been prescribed to many less probably somewhat less than 10M people. The Advair number seems very low to me since it has been around for 23 years and was able to continue patent protection beyond the expiration date not for the medication but the unique dispenser. 

I have the following observations based on what is available:

1.  Despite an overall increase in antidepressant prescriptions – new antidepressant prescribing and revenue is low.  This is probably due to the abundance of generic antidepressants and associated rationing of access to newer antidepressants by managed care companies (MCO) and pharmacy benefit managers (PBMs). My only interest in revenue is as a proxy for total number of prescriptions. Looking at the total revenue over years of availability in table 2 – antidepressants generated about $1-2B per year.

This is the downside to the antidepressant prescribing conspiracy theory.  If the business model was as simple as influencing mindless doctors there would not be such a drop in revenue.  There would be a never-ending revenue stream available just from that manipulation. The conspiracy theory fails to account for the regulatory model that rewards innovation with time limited patent protection but even then there are limits, in this case an abundance of generic drugs due to expired patent protection.

The conspiracy theory also fails to account for the fact that these patterns are well known and have happened in the past. When I was an intern one of my attendings commented that it was ironic that pharmaceutical companies were making billions off of indigestion with H-2 blockers while there was hardly any research being done on malaria - a disease that was killing a million people a year at that time. When the H-2 blockers (cimetidine, ranitidine, famotidine) became generics, proton pump inhibitors took their place in the billion-dollar profit cycle.  With the current regulatory landscape this cycle will continue to repeat.  At any given peak in the cycle it may appear there is a conspiracy to increase Pharm profits – but it appears those suggestions only happen when psychiatric medications are involved.

2.  The current top selling drugs take the approach of selling to very large populations or selling very expensive agents to smaller populations. Atorvastatin as an example generates about 3 times a much revenue per year as an antidepressant targeting about the same population size.  I have some insider information about atorvastatin and its manufacturing.  I was told by a chemist that at the time of its original manufacturing it was the most expensive tablet every produced and that cost was $1.00 per pill.  Today it is possible to get a 90 day supply of 10 mg tablets for $3.44 to $6.44 and the generic manufacturers are still making money.

3.  One of the rationales for regulatory protection of patent protection is the high cost of research.  Is there any evidence this applies to the current landscape of psychiatric medications?  At least one industry analysis expects slower than expected growth in “mental health therapies” but much greater growth in immunological therapies, cancer therapies and obesity therapy.  All would appear to be driven by current profits in these areas.  The current down cycle in psychiatric research may also explain the attempts to resuscitate hallucinogens and psychedelics for this application. There has long been the outcry that the only reason these drugs have not been adequately researched was their classification as Schedule I compounds – but the research so far has not been impressive.

4.  Given the limitations, the most striking number in the tables are the number of “antidepressant overprescribing” articles in the popular and professional literature.  That number (100K) exceeds the total number of overprescribing references in Table 1. by 25%.  There is not much granularity there but when taken in combination with statements seen in both the popular press and research literature it may confirm what has been known for a long time – psychiatry is by far the most criticized specialty whether that criticism is rational or not.

The cycles of feast and famine in the pharmaceutical industry are always the product of innovation, marketing, and politics just like any other capitalistic enterprise in the US. There is an opportunity to do immense good and we have just seen that happen with the vaccines and medications invented to stop a pandemic. Even the best possible treatments have side effects and sometimes very bad side effects. That is the expected cost of treating any diverse biological population. Psychiatrists and the medications they use are not better than that biological constraint – but they certainly are not worse.

 

George Dawson, MD, DFAPA

 

 Supplementary:  I would like nothing better than to have more accurate information about top selling drugs and who is prescribing them. If you have available sources send them to me or post the links here.  I can reanalyze the data if it becomes available. 

Graphics Credit: Many thanks to Rick Ziegler for allowing the post of his photo.

 

References:

 

1:  Brumely J. The 15 All-Time Best-Selling Prescription Drugs.  Kiplinger.  December 17, 2017

2:  Urquhart L. Top drugs and companies by sales in 2018. Nature Reviews Drug Discovery. 2019 Mar 12:NA-.

3:  Urquhart L. Top companies and drugs by sales in 2019. Nature Reviews Drug Discovery. 2020 Apr 1;19(4):228-9.

4:  Arrowsmith J. A decade of change. Nature Reviews Drug Discovery. 2012 Jan 1;11(1):17.

5:  Moorkens E, Godman B, Huys I, Hoxha I, et al. The Expiry of Humira® Market Exclusivity and the Entry of Adalimumab Biosimilars in Europe: An Overview of Pricing and National Policy Measures. Front Pharmacol. 2021 Jan 8;11:591134. doi: 10.3389/fphar.2020.591134. PMID: 33519450; PMCID: PMC7839249.

6:  Aitkin M, Kleinrock M.  Global Use of Medicines 2023: Outlook to 2027. January 2023.

 

ADHD - 28 Discussion Points

 

There was some of the usual controversy in the media today.  Is Attention Deficit~Hyperactivity Disorder over diagnosed or underdiagnosed?  The usual controversy contained the usual stories of how easy it is to get a diagnosis of ADHD in some places.  In some places it seems like just a matter of expense - a thousand dollar test battery. In other places there are people disabled by the condition who cannot get adequate treatment.  In the meantime there are international experts cranking out reams of papers on the importance of diagnosing and treating this condition in childhood. Occasionally an article shows up in the papiers about the cardiovascular safety of these medications. And in the New England Journal of Medicine there was a paper about a higher incidence of psychosis due to these medications.  Where does the reality lie?

I was fortunate enough to have worked at a substance use treatment center for about 12 years just prior to retiring. Only adults were treated at that facility. A significant number of them were diagnosed and treated as children. There were also a significant number of patients newly diagnosed as adults - some as old adults in their 60s and 70s. Whether or not ADHD can occur as a new diagnosis during adulthood is controversial and establishing a history consistent with childhood ADHD is problematic due to recall errors and biases. Secondary causes of ADHD in adults such as substance use problems and brain injuries increases in prevalence.  Although I am speculating, secondary causes seem a more likely cause of attentional symptoms in adults and therefore acquired ADHD without childhood ADHD if it does exist is an entirely different problem.

Prescribing stimulants to patients who may have stimulant use disorders is problematic for a number of reasons. Initially we had an administrative safeguard on the practice. Stimulant prescriptions could only be approved with a second opinion by another psychiatrist after reviewing the record. Eventually we had a core of psychiatrists who practiced the same way and the second opinion was no longer necessary.  Over the course of 12 years I developed these discussion points.  I think they are a good example of the minimum ground you need to cover in an evaluation for ADHD.  I typically had a 60-90 minute time frame to work with and could see people on a weekly basis for 30 minute follow ups. These evaluations were often controversial and resulted in collateral contacts, typically with a family member who was advocating for the stimulant prescription. 

A few basic points about ADHD and establishing the diagnosis. Like many psychiatric disorders there is no gold standard test.  Like some of the media discussions, I have been told that a person underwent days of testing before they were given the diagnosis of ADHD.  These are typical paper and pencil tests, but there have also been tests based on watching a computer screen and even crude EEG recordings. There are a few places that use very sophisticated brain imaging techniques. Unfortunately none of these methods can predict a clear diagnosis or safe and effective use of a medication that can reinforce its own use.  That leaves clinicians with diagnostic criteria and and a cut off based on functional status as a result of those symptoms.  That may not sound like much, but it eliminates a large pool of prospective ADHD patients who have no degree of impairment and those who are obviously interested in possible performance enhancement rather than ADHD treatment.  

Stimulant medications are highly abusable, as evidenced by several epidemics of use dating as far back as 1929. We are in the midst of a current epidemic.  For those reasons it is important not to add to the problem as either the individual or population levels. In my particular case, I was seeing patients who were all carefully screening for substance use and adequate toxicological screening. Since they voluntarily admitted themselves into a treatment center it was also more likely that they recognized the severity of the problem and were more open to treatment.  Even against that background - it is worth covering the above points.  Covering those points often involves repetition because of cognitive problems in detox or disagreement.

These are just a few health and safety considerations. My main concern in this area is that psychiatric treatments somehow have the reputation that they don't require medical attention. They are somehow isolated from the rest of the body. The person prescribing this medication needs to assess the total health status of the individual and determine if the medication prescribed is safe to use. Cardiac and neurological conditions are at the top of that list. I gave a blood pressure example because I have been impressed with how many people tell me that their blood pressure was not checked after a stimulant prescription or a stimulant was started despite diagnoses of uncontrolled hypertension, cardiovascular disease, cardiac arrhythmias or cerebrovascular disease.  These were typically new prescriptions in older adults with no prior history of ADHD.  

Coexisting psychiatric disorders are also problematic. Most have associated cognitive symptoms if they are inadequately treated. That is not a reason to diagnose ADHD or start a stimulant medication.  Typical symptoms that can be caused by stimulants are have to be recognized and the medication must be stopped if adjusting the dose is not helpful.

It is important to keep the range of biological heterogeneity in mind. Once you have narrowed down a population of people who most likely have ADHD, they will not all have a uniform response to medication.  They may not all want to take medication.  As adults many stopped taking ADHD medication and adapted to a work and lifestyle that works very well for them. That is a very suitable outcome for an initial assessment.  There is another group who want to try a verbal therapy for ADHD in some cases because they recognize they can no longer take stimulants because they were escalating the dose. That is also a suitable outcome for the assessment. In those people who have ADHD are want to take a medication, I think a non-stimulant medication like atomoxetine is a good place to start. In my experience it works very well.  Disagreement about stimulants, especially in people with a stimulant use disorder typically requires extended conversations with the patient and their family. A quality control initiative can provide very useful data for that conversation. I suggest that any clinic or clinician who prescribes stimulants collect outcome data on those prescriptions.  The key piece of data is a comparison of the relapse rate of those patients taking stimulants compared with patients treated with non-stimulants. Other data could be collected as well - like how long the prescriptions were refilled. There are rules about collecting that data depending on your practice setting.  Check those rules first.  Outcome data will be the best data on whether a correct decision was made about prescribing the stimulant.

I added the following slide based on polypharmacy considerations in the paper cited in reference 1.  This is a common clinical problem that needs to be approached rationally and that includes limit setting on the concept that every side effect or symptom needs to be addressed by a medication rather than a medication discontinuation, reduction, or substitution.  I always include a discussion of rare but serious side effects, synergistic side effects, drug interactions, interactions with comorbid medical problems and associated medications, and very serious interactions that could lead to hospitalization or death, like serotonin syndrome. 

I am going to end on a note about countertransference based on a disagreement I had at a conference about my methods. The speaker advocated for prescribing stimulants as a general operating practice for anyone with ADHD. When I confronted him about the problem of substance use he claimed his motivation was that he considered it his priority to "help" people and he thought that stimulants were the most helpful medication.

Whether or not a medication is helpful for any psychiatric disorder depends on a very careful assessment and clinical expertise that considers several dimensions including the potential risks and benefits for the patient and the incorporation of the patient preferences and values into the clinical decision making process.  In my evaluations, I try to sum all of that up in an informed consent discussion. In the area of ADHD evaluation and treatment, that covers a lot of ground and there is no simple uniform recommendation.

George Dawson, MD, DFAPA

Photo Credit:  Many thanks to my colleague Eduardo A. Colon, MD for allowing me to use his photos. 

The Semi-Random Pathway To A Psychiatrist

 

Mischaracterizations about how people come to see psychiatrists are an ongoing phenomenon in the media. It is one of many false premises used to build the idea that there is a monolithic psychiatry out there that is really an evil empire. That stands in contrast to psychiatrists going to work every day and seeing a full schedule of patients that was compiled without their knowledge or input. I will try to explain what really happens in this post.  My explanation is complicated by very little research done on this topic.

In my experience and the experience of the colleagues I have worked with, it is rare for a person with mental health problems to be seen by a psychiatrist without previously having seen another physician or mental health professional. That is true for both emergent and elective situations.  It follows that those patients have also been treated with medications and psychotherapy, often many different medications from different classes and different forms of psychotherapy.  The period of treatment before seeing a psychiatrist is typically quite long. As an example, in my last job as a consultant at a large substance use treatment facility – I was seeing people in their 30s and 40s who had an onset of a psychiatric disorder in their teenage years, gone through various treatments, and were seeing me as the first psychiatrist they had seen in their life. That could have been as long as 5-6 years of illness onset with no treatment followed by another 10-20 years of ineffective or sporadic treatment.

This is a hard problem to delineate because of all the factors involved. There may be a preference to not see a psychiatrist because of the connotation that psychiatry is equated with a more serious illness or the misconceptions about psychiatry that are actively promulgated by special interests. In some cases, primary care physicians consider themselves to be the right person to manage anxiety or depression.  The recent collaborative care initiative and physician rather than patient-based consultation may reinforce that idea. I have seen patients who were in treatment with every possible alternative medicine practitioner and getting treatments that could either not possibly help them or in some cases were detrimental. Some of my patients were in cults who claimed to be helping them until they were rescued by family members and brought to psychiatric treatment. Self help is a popular approach and some of it is very good but in other cases it can also be detrimental.  When bibliotherapy is used, the advice often includes instructions on sleep, diet, and exercise.  Many people resort to self-diagnosis and self-medication. Supplements are generally expensive and ineffective. Alcohol, stimulants, cannabis, benzodiazepines, psychedelics, and other intoxicants are uniformly ineffective and often amplify the original problem or create new problems.

In some cases, the person already has a very healthy lifestyle and given popular recommendations it is tempting to just amplify the health factors.  More protein, more sleep, and more exercise. It is very doubtful that any of those factors alone or in combination will address a serious mental health diagnosis.  I have treated world class athletes in the top 1% of all exercisers who got no relief at all from exercise.

Insurance limitations and rationing has been a significant factor in the last 40 years. A study done showed that referrals from managed care companies seldom result in real appointments with psychiatrists – in many cases because the insurance company claims that psychiatrists are in their network when they are not.  These are ghost networks or phantom panels (1).

Practice setting is also an important factor. During my 22 years of acute inpatient care – all of the people I saw were either taken to the emergency department or presented there themselves.  Most were untreated for years.  Some were overtreated with medication combinations and were seeing me because of side effects. It was more likely they had seen a psychiatrist because we had an outpatient department and our colleagues’ patients were hospitalized, but reviewing those histories it was clear they had been untreated for years before being seen in the outpatient department.

I first learned about the importance of the pathway to physician care in medical school.  When you first start seeing patients, there is an excessive focus on the form of the medical evaluation rather than the phenomenology. In other words – you want to carefully document all the history, symptoms, and findings that lead to a diagnosis and treatment plan. You are under the mistaken assumption that is all there is to being a physician.  At least until your first attending asks: “Well – what is the reason this patient sought treatment at this time? Why did they decide to come into the emergency room right now?”  At that point you start to realize that there are not a set number of pathways to care and patient presentations – there are millions.  Your job is to describe that unique pathway for every individual patient that you see – no matter what the problem is.

The pathway to psychiatric care in the United States is semi-random. People rarely make a conscious decision to see a particular psychiatrist and then see that person. At some level psychiatrists are the treatment providers of last resort and you are likely to see one in an emergency that you never planned or when another treatment provider tells you that it might be a good idea.  It is likely that a lot of disability and distress occurs in the meantime, along with a lot of bad advice. I think it is reasonable to try self help and other qualified mental health professionals first.  But if you are not seeing any good results – I would not let it go on for too long.  Like most things – if what you read in the papers about how a certain treatment for your problem seems too good – it probably is.

George Dawson, MD, DFAPA

Supplementary 1: 

Here is a post from about a year ago on involuntary treatment.  Look at the diagrams to see the number of personnel and steps involved in the process - apart from the psychiatrist. In a hospital setting - activating the civil commitment process typically occurs when an emergency department doctor or hospitalist places the person on an emergency hold.  In some states only police officers can initiate an emergency hold and in others any interested person.  An entire series of decisions not made by any psychiatrist determines if that person is eventually held or released.

Supplementary 2:

This is a graphic that I made from another application showing some alternate pathways to psychiatric treatment.  In almost all cases there is no psychiatric contact until the tier of treatments designated by white and green rectangles and possibly in the team approaches.

References:

1:  Malowney M, Keltz S, Fischer D, Boyd JW. Availability of outpatient care from psychiatrists: a simulated-patient study in three U.S. cities. Psychiatr Serv. 2015 Jan 1;66(1):94-6. doi: 10.1176/appi.ps.201400051. Epub 2014 Oct 31. PMID: 25322445.