Rearranging Deck Chairs While the Titanic Goes Down
ASCP deprescribing recommendations miss the point
Rearranging Deck Chairs While the Titanic Goes Down—The American Society of Clinical Psychopharmacology Issues Deprescribing Recommendations
In February, the American Society of Clinical Psychopharmacology (ASCP) issued a series of recommendations on psychotropic drug deprescribing.1 Derived from a Delphi survey completed by a 45-member international task force, the recommendations purport to offer guidance to clinicians on when and how psychiatric drugs should be deprescribed. Consensus was deemed to have been achieved if 75% of respondents endorsed a recommendation, and the task force reached agreement on 44 of 50 Delphi statements.
The paper includes welcome recognition of the fact that “few empirical studies have focused on optimal strategies to discontinue medication”, calling for better research to inform decision-making in this area. Equally welcome are the recommendations to routinely consider deprescribing when psychiatric drug treatment has not been effective or is not clinically indicated for treatment of the underlying condition.1 If consistently applied, these recommendations would represent a welcome deviation from current practice, where medications are routinely continued unnecessarily to the detriment of many patients. The authors of the ASCP recommendations rightly emphasize periodic risk–benefit reassessment, shared decision-making, and the need to consider patient preferences and cultural context. These principles represent meaningful progress.
Overall, however, the ASCP recommendations highlight concerning gaps. An unintended consequence of the Delphi process is to provide a snapshot of the state of deprescribing knowledge among psychopharmacology professionals. The ASCP recognises this itself when it acknowledges that the recommendations are primarily based on ‘intuitive knowledge’ rather than objective evidence. Forty years after most of these drugs were first brought to market, we might hope that professionals would by now have a little more to go on than ‘intuition’. That there have still been so few studies on how best to stop medications is damning in itself.
It is even more unfortunate that much of this intuition is incorrect, reflecting fundamental misunderstandings about the risks and benefits of discontinuing psychiatric drugs, and an ongoing reliance on outdated models for understanding psychiatric drug withdrawal. Clinicians seeking practical guidance about how to deprescribe in a safe, pharmacologically rational manner will find little here to assist them.
More fundamentally, the ASCP recommendations repeatedly mistake the forest for the trees. The document approaches deprescribing largely as a matter of refining existing prescribing practice, rather than confronting the broader reality that psychiatry has normalised long-term psychotropic use while investing remarkably little effort into understanding its long-term consequences. The imbalance in emphasis is striking. The recommendations devote space to highly specific edge cases, such as whether SSRIs should be deprescribed in older adults taking thiazide diuretics because of possible hyponatraemia risk, while offering little practical discussion of the problems faced by the far larger population of long-term antidepressant users attempting to discontinue treatment. The result is a document focused on optimising the margins of prescribing practice while still sidestepping the central clinical and public health issues raised by widespread long-term psychiatric drug use.
It is also troubling to read the ASCP’s ill-considered comments about the hypothesised relationship between deprescribing and ‘antipsychiatry’, which imply that there may be something inherently suspicious about seeking to deprescribe at all. Adopting a position of reflexive hostility to even mild critiques of current practice, the recommendations continually cast doubt on the credibility of patients reporting adverse effects or lack of efficacy.
Misunderstanding the causes of withdrawal
Although incidence rates are debated, around half of patients who discontinue psychiatric drugs after more than a few months of use will experience withdrawal symptoms. The figure is higher for longer term users, and withdrawal can be both severe and long lasting for a substantial number of people.2–4 Recent papers, relying on studies that did not set out to evaluate withdrawal and examining only people who have taken these drugs for short periods, have underestimated the risk of withdrawal. These papers cannot be extrapolated to the average person taking these drugs—short term users are in the minority and therefore unrepresentative of the population at large.5–7
Despite increasing recognition of these issues, the ASCP recommendations do not directly address withdrawal as a risk associated with psychiatric drug discontinuation. Where withdrawal is indirectly mentioned, the views of the task force reflect a misunderstanding of the pathophysiological processes that cause withdrawal symptoms. One such recommendation suggests that drugs with long half-lives, such as fluoxetine, are ‘auto-tapering’ and can therefore be safely stopped abruptly without gradual dose reduction.1 Another, which narrowly failed to reach consensus, states that ‘discontinuation phenomena’ (a euphemistic term for withdrawal symptoms) can be relieved by switching to a drug with a longer half-life.
These recommendations reflect the widespread belief that withdrawal symptoms are mediated simply by the time taken for the drug to be eliminated from the body. This belief would suggest that as soon as the drug has left the body withdrawal symptoms should cease—a clearly absurd proposition to anyone who has experienced withdrawal effects for more than a few days after stopping a drug.
In reality, withdrawal effects are caused by a mismatch between what the brain and body have come to expect during exposure to the drug and the reduced input when the drug is reduced in dose.4 The key fact that has been misunderstood by the mainstream is that adaptations caused by antidepressants and other psychiatric drugs have been shown to persist for months and years after drug discontinuation, long after even long half-life drugs have been eliminated.4 When a drug is reduced or discontinued, these adaptations are unopposed. Patients experience withdrawal symptoms for as long as it takes these adaptations to reverse and for the system to return to a pre-drug state (its ‘factory settings’).
An analogy may be helpful: when you go to a loud concert your eardrums become less sensitive to sound to maintain homeostasis. When you go out into the quiet street afterwards your companions’ voices sound muffled. This is a sound withdrawal syndrome. The reason it takes a minute or two before your companions’ sound normal is because that is how long it takes your eardrum to relax—despite the fact that the sound dissipates the instant you shut door of the concert. In other words, it is the time taken for your body’s adaptation to a stimulus to dissipate that explains how long a withdrawal syndrome persists; not just the time for the stimulus to cease.
For this reason, it is not correct to characterise longer half-life drugs such as fluoxetine as ‘auto-tapering’. Although drugs with longer half-lives do have lesser withdrawal risk than other similar drugs, they still regularly cause withdrawal symptoms. For example, double-blind randomised control trials show that 50% of people discontinuing fluoxetine experience withdrawal symptoms.4 Withdrawal risk from fluoxetine is only moderately lower than for other drugs in the same class.8
The ASCP fails to recognise that half-life only has moderate effect on withdrawal risk because the time for elimination is still short compared to the longer timeline for adaptations to reverse. Drugs with longer half-lives are particularly likely to cause delayed onset of withdrawal symptoms. Clinicians often misdiagnose these presentations, precisely because they continue to believe the misconceptions about drug half-life and withdrawal that the ASCP recommendations perpetuate. Clinical vigilance should be maintained when long half-life drugs are being discontinued, with monitoring continuing over a longer period to allow for detection of delayed-onset symptoms. Such drugs still need to be carefully tapered to prevent withdrawal effects. It is therefore ironic to see Goldberg, the lead author of the ASCP recommendations, quoted in The New York Times saying that advice to taper all drugs carefully and slowly is ‘unscientific’.9
Relapse or withdrawal?
It now seems probable that withdrawal, rather than relapse, is the principal risk associated with psychiatric drug discontinuation, and this is reflected in emerging research. Common emotional withdrawal symptoms, such as low mood and anxiety, overlap with the symptoms of many mental health conditions and have generally been misclassified as relapse in discontinuation studies.10 This conflation of withdrawal and relapse casts doubt on the assumed relapse prevention properties of many psychiatric drugs. Discontinuation studies used to gauge relapse prevention properties generally find an excess of about 20% relapse in the group which stops medication compared with the group that continues.11 Given that withdrawal occurs in about half of people who stop medication it is likely that misclassification of withdrawal effects explains most if not all of this difference in apparent relapse rates. This holds true even if lower estimates for withdrawal effects are relied on. Indeed, even the idea that mental health conditions are lifelong illnesses which inevitably relapse without treatment is shakily supported. Even so, the ASCP recommendations mistakenly position relapse as the main risk associated with deprescribing.
This has important consequences for the tone of the ASCP recommendations, which seem to envision few circumstances in which the benefits of discontinuation should be seen as outweighing the assumed risk of relapse. The recommendations are unnecessarily conservative about when it is appropriate to deprescribe, and this conservatism may logically be connected to the long list of pharmaceutical company conflicts of interest declared by the authors. People who work closely with the companies who sell psychiatric drugs are unlikely to recommend widespread deprescribing of their sponsors’ products.
Reinstatement, extended drug treatment or indeed discouraging patients from attempting discontinuation in the first place are the logical mitigations of relapse risk. By contrast, the risk of withdrawal can be ameliorated by slow, personalised, hyperbolic tapering, enabling patients to successfully discontinue drugs that are causing adverse effects, are not clinically indicated or are simply unwanted.12 Although it is increasingly recognised in guidelines around the world,13 the ASCP recommendations do not mention the central role of hyperbolic tapering in increasing the likelihood of good patient outcomes.14,15 Nor do they outline the practical mechanics of how gradual tapering might be achieved. Clinicians are vaguely advised to monitor patients closely, without guidance about what to do if difficulties are encountered.
Much is made throughout the recommendations about the need to undertake “collaborative risk-benefit assessments with patients” and otherwise engage in shared decision-making. Yet the ASCP nowhere advises clinicians to discuss withdrawal symptoms and their management with patients, including how often people mistake withdrawal effects for a return of their condition, trapping them in pessimistic expectations. Receiving clear, accurate information about the drug discontinuation process reduces anticipatory anxiety, and has been identified as a key facilitator of successful deprescribing.16 The fatalistic and risk-averse narrative promoted by the ASCP would no doubt have the opposite effect. A more useful and accurate approach would emphasise that withdrawal symptoms are highly manageable when tapering is conducted correctly.
Is deprescribing a dog whistle for ‘antipsychiatry’?
The first matter which the ASCP task force voted on, and presumably therefore the issue it considered most important, concerned the political connotations of the term ‘deprescribing’ itself. According to the commentary accompanying the recommendations, ‘deprescribing’ has become unacceptably tainted by association with “the antipsychiatry community”. The authors of the ASCP recommendations expand on this view in an explanatory paper, where they outline the ‘sociolinguistic controversy’ that is said to be attached to the word ‘deprescribing’.
‘Antipsychiatry’ is defined so broadly in this paper that it would appear to apply to anyone with any doubts at all about psychiatric drug efficacy or safety. Modest, evidence-based concerns about drug toxicity and overprescribing are characterised as ‘unwarranted accusations’ against the psychiatric profession. While the ASCP ultimately decided to retain the word ‘deprescribing’, it is not encouraging to see legitimate critiques of the current psychiatric paradigm so comprehensively dismissed and delegitimised. Improving patient health and reducing unnecessary harm are at the heart of these critiques. It also reveals the source of the defensiveness which animates this document, preventing leading psychiatrists from thinking through issues of public health and safety more soberly.
Patients as unreliable witnesses
This unwillingness to acknowledge critique extends to a deep suspicion of patients reporting adverse effects or lack of efficacy from psychiatric drug treatment. It almost seems that the ASCP regard patients reporting such experiences as ‘antipsychiatry’. Much is made of the need to ensure ‘adequate adherence’ before accepting a patient’s word that they have not found a drug efficacious. Lack of efficacy is almost always assumed to be an individual failure on the part of the patient, rather than a sign that the drug is simply not effective. What non-adherence may in itself signify about tolerability, the burden of adverse drug effects or the personal preferences and values of the patient is not explored.
Indeed, despite the emphasis on shared decision-making, the recommendations are dismissive of patient preference and experience. Adverse effects like emotional blunting, sexual dysfunction and weight gain, which can seriously degrade a patient’s quality of life, are described in the recommendations as merely ‘bothersome’. The emphasis is on ‘managing’ these effects, often by the addition of new drugs. For example, GLP-1 agonists for weight gain, or bupropion for sexual dysfunction. A rational desire to discontinue medication to avoid drug dependence, or because non-pharmacological treatments are logically preferred because they have fewer adverse effects, is nowhere countenanced as legitimate.
This is particularly striking in relation to the ASCP recommendations on deprescribing in pregnancy. Legitimate concerns about taking psychiatric drugs during pregnancy are dismissed by the recommendations as, “ill-informed” and “not…evidence-based”. Although the full impact of foetal drug exposure remains contested, there is abundant evidence suggesting that drugs like antidepressants have wide-ranging effects on foetal development. Alongside this, much of the evidence for the benefits of drugs like antidepressants during pregnancy stems from the methodologically suspect discontinuation studies addressed above. In this context of uncertain benefits and clear harms, it is perfectly rational for women to wish to minimise exposure.
The recommendations also risk perpetuating the gaslighting of patients by interpreting difficulties stopping medications as primarily psychological problems, rather than withdrawal effects that are the predictable physiological consequence of physical dependence. The consensus statement reframes such difficulties through speculative psychodynamic concepts such as “attachment styles”, medications functioning as “transitional objects”, or unconscious fears about loss of care and validation. Of course such dynamics may exist in a minority of cases, as they might in any area of medicine. But foregrounding these ideas in a consensus statement on deprescribing risks implying that problems on stopping mainly reflect emotional dependency or distorted beliefs, rather than concerns grounded in lived experience and increasing evidence. A patient who fears becoming destabilised after stopping an antidepressant may not be expressing unconscious dependency needs; they may simply have experienced severe withdrawal symptoms before.
Conclusion
Overall, it is a good sign that the psychiatric establishment is reckoning with deprescribing. Amplified by the interest of the current US administration, public outcry has become too loud to ignore. However, the enthusiastic welcome granted this woefully inadequate document by media such as The New York Times shows just how low the bar is. Should the psychopharmacology profession really be celebrated for recognising the need for safe deprescribing only now, forty years after the drugs were first released and only now that countless patients have been harmed? Did the tobacco industry deserve praise for recognising the adverse effects of smoking only after decades of misinformation campaigns? Would we cheer car manufacturers for suggesting the brake pedal might be installed for some models only after forty years of catastrophic traffic accidents? It is not clear that the same group who oversaw the mass over-prescribing of psychiatric drugs is best placed to address a problem that they themselves have caused.
This broader concern is compounded by the practical failures of the consensus statement, and the real world consequences of its dismissive attitude to patients. The recommendations ignore the overestimation of long term-benefits derived from flawed discontinuation trials, and sidestep the major issue of withdrawal. The practical mechanics of how drug discontinuation should be safely managed, as outlined in detail in places like the Maudsley Deprescribing Guidelines,12 are entirely absent. Most concerningly, patients burdened by adverse effects or concerned about drug dependency are likely to continue to be dismissed, and to discontinue their medication without clinical oversight. Peer to peer tapering communities have filled the vacuum of safe deprescribing knowledge left by organisations like the ASCP. The profession would do well to learn from their hard-earned experience, rather than seeking to undermine their credibility in order to protect its own faltering reputation.
This piece was first published by Mad in America. Check it out on their website, and give their Substack a follow:
References
1 Goldberg JF, Swartz HA, McIntyre RS, et al. The American Society of Clinical Psychopharmacology (ASCP) task force on the deprescribing of psychotropic medications for mood disorders: Delphi expert consensus. Br J Psychiatry 2026; : 1–9.
2 Davies J, Read J. A systematic review into the incidence, severity and duration of antidepressant withdrawal effects: Are guidelines evidence-based? Addict Behav 2019; 97: 111–21.
3 Horowitz M, Buckman J, Saunders R, Aguirre E, Davies J, Moncrieff J. Antidepressants withdrawal effects and duration of use: a survey of patients enrolled in primary care psychotherapy services. Psychiatry Res 2025; : 116497.
4 Horowitz M, Framer A, Hengartner MP, Sørensen A, Taylor D. Estimating risk of antidepressant withdrawal from a review of published data. CNS Drugs 2023; 37: 143–57.
5 Kalfas M, Tsapekos D, Butler M, et al. Incidence and nature of antidepressant discontinuation symptoms: A systematic review and meta-analysis: A systematic review and meta-analysis. JAMA Psychiatry 2025; published online July 9. DOI:10.1001/jamapsychiatry.2025.1362.
6 Henssler J, Schmidt Y, Schmidt U, Schwarzer G, Bschor T, Baethge C. Incidence of antidepressant discontinuation symptoms: a systematic review and meta-analysis. Lancet Psychiatry 2024; 11: 526–35.
7 Moncrieff J, Hobday H, Sørensen A, et al. Evidence on antidepressant withdrawal: an appraisal and reanalysis of a recent systematic review. 2025; 55: e191.
8 Gastaldon C, Schoretsanitis G, Arzenton E, et al. Withdrawal Syndrome Following Discontinuation of 28 Antidepressants: Pharmacovigilance Analysis of 31,688 Reports from the WHO Spontaneous Reporting Database. Drug Saf 2022; 45: 1539–49.
9 Barry E. Top Psychiatrists Call for a Greater Focus on Ceasing Medication. The New York Times. 2026; published online May 1. https://www.nytimes.com/2026/05/01/science/psychiatry-kennedy-ssris-maha-antidepressants.html (accessed May 4, 2026).
10 Horowitz MA, Taylor D. Distinguishing relapse from antidepressant withdrawal: clinical practice and antidepressant discontinuation studies. BJPsych Advances 2022; 28: 297–311.
11 Hengartner MP. How effective are antidepressants for depression over the long term? A critical review of relapse prevention trials and the issue of withdrawal confounding. Therapeutic Advances in Psychopharmacology 2020; 10: 2045125320921694.
12 Horowitz M, Taylor D. The Maudsley Deprescribing Guidelines in Psychiatry: Antidepressants, Benzodiazepines, Gabapentinoids and Z-drugs. New York, USA: Wiley-Blackwell, 2024 https://scholar.google.com/citations?view_op=view_citation&hl=en&citation_for_view=sEcbBuUAAAAJ:abG-DnoFyZgC.
13 Horowitz M, Wilcock M. Hyperbolic tapering of antidepressants: where are we now? Drug Ther Bull 2025; 64: 3–7.
14 Cooper RE, Ashman M, Lomani J, et al. “Stabilise-reduce, stabilise-reduce”: A survey of the common practices of deprescribing services and recommendations for future services. PLoS One 2023; 18: e0282988.
15 van Os J, Groot PC. Outcomes of hyperbolic tapering of antidepressants. Ther Adv Psychopharmacol 2023; 13: 20451253231171520.
16 Maund E, Dewar-Haggart R, Williams S, et al. Barriers and facilitators to discontinuing antidepressant use: A systematic review and thematic synthesis. J Affect Disord 2019; 245: 38–62.
The ASCP is making sure they are seen showing-up to the deprescribing conversation, deliberately deflecting accountability and concern for patient safety in favor of their transparent panic around the failing reputation of psychiatry overall.
So well said!