How psychiatric drug marketing feeds itself

Nobody in this story has to be a villain. a drug is approved on a short trial its maker ran. the maker pays a few dozen psychiatrists to teach the rest. prescriptions follow. the side effects create a market for the next drug. and the system that would notice a long-term problem catches a small fraction of what happens. every step is legal, most of it is disclosed, and each one makes the next more likely. this page walks the loop with the federal numbers, says where the evidence is weaker than the story, and ends with what it means if you are the one taking the pill.

nothing here is a reason to start, stop or change a medication. that is a decision to make with your prescriber. stopping suddenly carries its own risks.

1. the drug arrives with a short trial its maker ran

the median trial behind an antidepressant approval lasted eight weeks (every drug, every trial length). the company designs it, pays for it and decides what gets published. when Turner compared the FDA’s files with the journals, 94 percent of published antidepressant trials looked positive; in the FDA’s complete set, 51 percent were. Cochrane’s review of sponsorship found industry-funded studies reach favourable conclusions more often (risk ratio 1.27) with methods that look just as clean. the agency that reviews the application is itself 73 percent funded by the applicants’ fees. funded by is not controlled by. it is the first link.

2. the maker pays a few dozen psychiatrists to teach the rest

in 2024 companies reported $71.4 million in payments to US psychiatrists. most of those payments were a meal. more than half the money, 56 percent, was speaker fees: a physician paid to deliver a company-written presentation to other prescribers. sixty psychiatrists received $28.3 million of the total. ten companies paid 94 percent of that, and every product attached to it was a brand still under patent. no generic appears (the list, from the federal file).

the government’s own fraud watchdog has said what it thinks this is. the HHS Inspector General, 2020: “OIG is skeptical about the educational value of such programs”, because the same information is free in the label and the journals, which “further suggests that at least one purpose of remuneration associated with speaker programs is often to induce or reward referrals.” it also described what past cases found: companies that “selected high-prescribing HCPs to be speakers and rewarded them with lucrative speaker deals”.

3. what the paid psychiatrists prescribe

we joined the two federal files by prescriber number: the sixty best-paid psychiatrists of 2024, and what each wrote for Medicare patients the same year. the comparison group is every psychiatrist in the Medicare prescribing file, 23,932 of them. the products are the twenty brands the payments were attached to.

• for all psychiatrists, those twenty brands were 2.3% of prescriptions and 54% of the cost: $2.17 billion of $4.03 billion. the file hides any drug a prescriber wrote fewer than eleven times, so the true share is a little higher. more than half of psychiatrists had no countable prescribing of any of them.
• for the 52 of the sixty who appear in the file, the same brands were 6.5% of prescriptions. that is 2.8 times the rate of their peers, and it does not change when the single largest prescriber is removed (6.4%). 25 of the 52 were in the top tenth of all psychiatrists on this measure.
• and 7 of the fourteen best paid barely prescribe in Medicare at all: they are missing from the file or wrote fewer than 100 prescriptions in it. the median psychiatrist wrote 494.

read those together. the very top of the list is not being paid for its own prescription pad. those are national speakers and authors whose value is their reach into everyone else’s. below them is a second tier of high-volume community prescribers who use the promoted brands far more than their peers and are paid to talk about them.

what this does not show. it does not show that a payment caused a prescription. companies pick speakers who already like the drug, which the Inspector General says in so many words. some of these clinicians run long-acting injectable clinics or treat the most treatment-resistant patients, where newer brands are a reasonable choice. Medicare is one payer, and the file hides any drug a prescriber wrote fewer than eleven times. across all of medicine, a 2021 review of 36 studies found payments associated with more prescribing of the paying company’s drug in every study, with evidence of timing in nine and of dose-response in twenty five. the authors call that suggestive of cause. it is still observational. we name no individual here because a group pattern is not evidence about any one person’s motives.

4. the side effect becomes the next product

this is where the loop closes. tardive dyskinesia is an involuntary movement disorder caused by antipsychotics. a meta-analysis of 41 studies put its prevalence at 20.7 percent in people currently on the newer, second-generation drugs, and at 7.2 percent in those who had never taken an older one. two of the most heavily promoted products in the payment file, Ingrezza and Austedo, exist to treat it. in 2024 psychiatrists alone wrote 84,402 Medicare prescriptions for them at a cost of $647 million. the same speaker circuit that promotes the newest antipsychotics promotes the drugs for the movement disorder antipsychotics cause, and many of the same people are paid for both.

the mitigating side: these are real treatments for a real and disabling condition, tardive dyskinesia was worse on the older drugs (30.0 percent), and nobody designed the sequence. but a system paid by the prescription has no financial reason to prefer the lowest dose, the shortest course or the older generic, and every reason to treat each new problem with a new brand.

5. what nobody measures

the loop would be self-correcting if long-term harm were counted. mostly it is not.

• the trials are weeks; the use is years. eight weeks at the median for an antidepressant approval, against about five years of real-world use, and one user in four past ten years.
• safety problems surface late, and psychiatric drugs lead. of 222 drugs approved from 2001 to 2010, 32 percent later had a withdrawal, a new boxed warning or an FDA safety communication. the first came a median of 4.2 years after approval. drugs for psychiatric disease had these events at 3.78 times the rate of other drugs.
• the follow-up studies often never happen. for drugs approved on limited evidence from 2005 to 2012, no controlled study at all was published after approval for 35 percent of the indications.
• the alarm system is voluntary and nearly silent. across 37 studies in 12 countries, the median share of adverse drug reactions that were never reported was 94 percent. for serious reactions the medians still ran from 80 to 95 percent depending on the setting.

put that beside the outcomes: more people in treatment every year, no fall in the population measures. a market that grows whether or not people get better, with an alarm that catches six reactions in a hundred, will not notice on its own if something is going wrong over ten or twenty years. that is not a claim that something is. it is a claim that nobody is positioned to know.

what to do if you are the one taking the pill

• look up your prescriber in Open Payments. a payment is a fact, not a verdict. it is a fair thing to ask about.
• if you are offered a brand that launched in the last few years, ask what the older generic alternative is and why it is not the first choice for you.
• ask what the longest trial of your drug was, and what is known beyond it.
• report side effects yourself. FDA MedWatch takes reports from patients. almost nobody files one, which is part of why so little is known.
• never stop or cut a dose on your own. the library has the taper evidence.

Sources

1. CMS. Medicare Part D Prescribers, by Provider and by Provider and Drug, data year 2024. Joined to Open Payments 2024 by NPI; our arithmetic, 2026-09-18. https://data.cms.gov/provider-summary-by-type-of-service/medicare-part-d-prescribers
2. CMS. Open Payments Program Year 2024 General Payment Data. https://www.cms.gov/priorities/key-initiatives/open-payments/data
3. HHS Office of Inspector General. Special Fraud Alert: Speaker Programs. 16 November 2020. https://oig.hhs.gov/documents/special-fraud-alerts/865/SpecialFraudAlertSpeakerPrograms.pdf
4. Mitchell AP, Trivedi NU, Gennarelli RL, et al. Are Financial Payments From the Pharmaceutical Industry Associated With Physician Prescribing? A Systematic Review. Ann Intern Med 2021;174(3):353–361. https://doi.org/10.7326/M20-5665
5. Lundh A, Lexchin J, Mintzes B, Schroll JB, Bero L. Industry sponsorship and research outcome. Cochrane Database Syst Rev 2017;2:MR000033. https://doi.org/10.1002/14651858.MR000033.pub3
6. Turner EH, Matthews AM, Linardatos E, Tell RA, Rosenthal R. Selective publication of antidepressant trials and its influence on apparent efficacy. N Engl J Med 2008;358(3):252–260. https://doi.org/10.1056/NEJMsa065779
7. Carbon M, Hsieh CH, Kane JM, Correll CU. Tardive Dyskinesia Prevalence in the Period of Second-Generation Antipsychotic Use: A Meta-Analysis. J Clin Psychiatry 2017;78(3):e264–e278. https://doi.org/10.4088/JCP.16r10832
8. Downing NS, Shah ND, Aminawung JA, et al. Postmarket Safety Events Among Novel Therapeutics Approved by the US Food and Drug Administration Between 2001 and 2010. JAMA 2017;317(18):1854–1863. https://doi.org/10.1001/jama.2017.5150
9. Hazell L, Shakir SA. Under-reporting of adverse drug reactions: a systematic review. Drug Saf 2006;29(5):385–396. https://doi.org/10.2165/00002018-200629050-00003
10. Pease AM, Krumholz HM, Downing NS, Aminawung JA, Shah ND, Ross JS. Postapproval studies of drugs initially approved by the FDA on the basis of limited evidence: systematic review. BMJ 2017;357:j1680. https://doi.org/10.1136/bmj.j1680
11. FDA. MedWatch: The FDA Safety Information and Adverse Event Reporting Program (voluntary reporting by consumers and health professionals). https://www.fda.gov/safety/medwatch-fda-safety-information-and-adverse-event-reporting-program

adam