New Deprescribing Guidance Marks an ‘Important Moment’ in Psychiatry

A new expert consensus statement from the American Society of Clinical Psychopharmacology offered clinical guidance on when and whether to discontinue psychiatric medications. The statement, developed by a 45-member international task force, reached consensus on 44 of 50 recommendations addressing when deprescribing is warranted.

Among the key points: clinicians should periodically reassess all medications, always verify adherence before concluding a drug isn’t working, and engage patients in shared decision-making about discontinuation. The recommendations address what authors said is a long-neglected aspect of psychopharmacology practice.

“Prescribers in mental health are often taught how to start medicines but not how to end them, or how to recognize an endpoint,” lead author Joseph F. Goldberg, MD, MSc, clinical professor of psychiatry at the Icahn School of Medicine at Mount Sinai, New York City, told Medscape Medical News.

The study was published online on February 25 in JAMA Network Open.

Much-Needed Guidance

Psychiatric training and research have historically focused on selecting, dosing, and combining medications, with far less attention paid to when and how to stop psychotropic drugs. That has left many patients on complex regimens without clear guidance on treatment duration or regular reassessments to determine whether a medication should be continued.

The task force sought to address this disconnect by establishing principles for deciding whether and when to stop a medication safely, distinguishing from prior literature that focused mainly on the mechanics of tapering off a drug.

“There’s been a lot in the popular media about the term deprescribing. It’s been seized by certain groups to mean inappropriateness of medicines in psychopharmacology the idea that medicines are hazardous and dangerous and shouldn’t be used,” Goldberg said. “We were very alarmed by the lay public messaging. We wanted to say some things for the professional record,” he added.

For the project, panelists conducted a review of existing literature and developed and completed a Delphi survey between January and May of 2025. Each survey item underwent multiple rounds of review, and consensus on each statement was defined as agreement or disagreement by 75% or more of the panelists.

Key Recommendations

Several recommendations achieved unanimous agreement. Panelists agreed that the usefulness of every psychotropic medication should be reassessed periodically at minimum, annually and that a risk-benefit analysis should be conducted before any deprescribing decision. There was also unanimity that when patients take multiple psychiatric medications, only one deprescribing change should be made at a time.

“The most fundamental takeaway is to not passively re-prescribe and renew without periodically stopping to assess the regimen,” Goldberg noted.

The task force strongly endorsed assessing medication adherence before drawing conclusions about efficacy. Given that approximately 50% of patients with chronic mental health conditions don’t adhere well to a treatment regimen, the panelists agreed that decisions about deprescribing for perceived lack of efficacy shouldn’t be made until the patient’s adherence has been carefully evaluated.

“All clinicians in psychiatry understand and recognize that adherence is a tremendous problem it’s arguably the leading cause of relapse in mood disorders and in schizophrenia,” said Goldberg. “But I think most don’t know how to assess it.” Clear communication about the rationale for deprescribing achieved 98% consensus.

As to benzodiazepine use among specific populations, the task force reached 100% agreement that it should be carefully reassessed as patients become older adults, and that medication lists in older patients should be routinely evaluated for cumulative anticholinergic burden.

Areas of Disagreement

Not all recommendations achieved consensus. When asked whether valproate should be routinely deprescribed in women of childbearing potential regardless of whether pregnancy is planned only 67% of panelists agreed, falling short of the 75% threshold. This was notable given valproate’s well documented teratogenic risks and a World Health Organization (WHO) statement advising against its use in that population. Goldberg acknowledged the finding was surprising.

“You never say never,” he said. “There may be circumstances where the pros do outweigh the cons. Some people thought the WHO recommendation was taking it a bit far they’d go along with avoiding it in pregnant women or sexually active women, but in terms of all women of reproductive potential, it may be going a bit far.”

The task force also did not reach consensus on replacing anticholinergic medications used to manage antipsychotic-induced movement disorders with alternatives such as amantadine (68% agreement). And panelists were divided on the common clinical practice of not restarting a medication that a patient previously overdosed on; only 46% agreed on that practice after two rounds of discussion.

Among the most common concerns raised by panelists was the challenge of developing general guidance for clinical situations that are often patient specific. The lack of empirical studies on deprescribing also led panelists to suggest that some recommendations may be more intuitive than evidence based.

“This supports the third-most common concern of panelists: because most serious psychiatric conditions are highly recurrent or chronic, true pharmacological endpoints are often elusive,” the authors wrote.

“Panelists often eschewed all-or-nothing generalizations about deprescribing in the absence of more nuanced information about a given patient’s circumstances, clinical status, and preferences, instead advocating for a dialogue with patients as stakeholders in their own outcomes,” they added.

An ‘Important Moment’

Experts have long acknowledged the lack of consensus in the field on whether or when to stop psychotropic medication. That makes the development of the statement “a very important moment in psychiatry,” said Allen Frances, MD, professor emeritus of psychiatry and behavioral sciences at Duke University School of Medicine, Durham, North Carolina, and chair of the task force that developed the Diagnostic and Statistical Manual of Mental Disorders, fourth edition.

“Clinicians are trained from the very first minute they begin psychiatric residency to prescribe medicines. They’re very poorly or not trained at all in stopping [deprescribing] them,” Frances told Medscape Medical News. “Any idiot can prescribe medicine. It takes real skill to deprescribe medicine.” The adherence problem is compounded by poor prescribing practices, he added.

“Very often, people were never given an adequate dose of the first medicine. It’s never determined for sure whether it worked, and rather than raise the dose, another medicine is added,” said Frances, who was not part of the task force that developed the guidance. “The overall result is careless, reckless polypharmacy.” Authors of an accompanying editorial agreed.

“The relative lack of attention to deprescribing has often left individuals without clear guidance about the anticipated duration of treatment, while saddling some with unnecessary adverse effects, drug interactions, and costs that might have been averted with more critical reappraisal of their medications at regular intervals,” corresponding author Jonathan E. Alpert, MD, PhD, Department of Psychiatry and Behavioral Sciences at the Albert Einstein College of Medicine, Montefiore Einstein, Bronx, New York, and colleagues wrote.

“The apparent lack of priority accorded to deprescribing in psychiatric practice has also reinforced, in some quarters, mistrust of the field as being unduly biased toward long-term pharmacotherapy or overly complacent about refilling prescriptions beyond the point they are needed,” the editorialists wrote.

The consensus statement brings the issue of stopping psychiatric medication to the foreground of clinical practice and study, marking “an important milestone in psychopharmacology,” the editorialists added. While many of the recommendations may not be new to long-time clinicians, the authors said that areas that achieved consensus are “worth integrating into best practices.”

They also took note of topics that were met with disagreement, including that only two thirds of panelists agreed with routinely discontinuing valproate in women of childbearing potential, adding that “the use of valproate among women of childbearing potential is widely discouraged, including by the World Health Organization and by most reproductive psychiatrists due to definitive evidence for teratogenicity.”

They also took particular note of the lack of consensus among task force members on the common practice of restarting a patient on a medication involved in a previous overdose was of particular note. The lack of consensus “likely reflects the recognition that this prevalent practice is based on clinical lore rather than research,” they wrote.

Note: This article originally appeared on Medscape.