Every Medication Needs a Current Job Description

by | Sep 17, 2026 | Medication Decisions | 0 comments

A psychiatric medication can be appropriate to start and still deserve a serious reassessment later.

That should not be a controversial statement. It is simply how good decisions work when circumstances change.

In February 2026, a 45-member international task force convened by the American Society of Clinical Psychopharmacology published a consensus statement on psychotropic deprescribing. Its clearest point was not that people should come off psychiatric medication. It was that every psychiatric medication deserves periodic reassessment, at minimum annually. ([jamanetwork.com](https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2845497?utm_source=openai))

I keep coming back to the implication: every medication needs a current job description.

Not the job it had five years ago. Not the diagnosis attached to the original prescription. Not the assumption that a refill means the question has already been answered.

The starting question is not “Should I stop?”

Medication conversations often become binary too quickly.

Continue, or discontinue. Helpful, or harmful. Relapse, or withdrawal.

Those categories can be important, but they are often too blunt to describe the decision in front of someone.

A better first question is: What specific benefit is this medication expected to provide now?

The answer may be clear. Perhaps it has reduced a symptom that was severe, recurrent, and disabling. Perhaps it has made sleep more reliable, panic less frequent, mood more stable, or daily functioning more possible. Continued treatment may have real value, and that value should not be minimized simply because the medication has been used for a long time.

But sometimes the answer is less clear. The medication may have been started during an acute period that has since passed. It may have helped initially but never been revisited. Another medication may have been added to manage a problem that emerged after the first one. The person may no longer know which medication is meant to address which symptom.

That is not a failure of character or compliance. It is what can happen when treatment accumulates faster than it is reviewed.

Reassessment is not a verdict

The new consensus statement is useful precisely because it frames deprescribing as a risk-benefit decision, not as a moral achievement. The group identified several reasons a medication might warrant consideration for removal: lack of meaningful benefit, adverse effects that outweigh benefit, redundant medications, interactions, changed diagnosis, or changed treatment goals. It also emphasized patient participation and the need for closer monitoring when a change is made. ([doi.org](https://doi.org/10.1001/jamanetworkopen.2026.0043?utm_source=openai))

Here is the part worth looking at: this was a Delphi consensus process and a selective literature review. It was not a clinical trial proving that annual medication reviews produce better long-term outcomes. The authors explicitly noted that implementation studies are still needed. ([doi.org](https://doi.org/10.1001/jamanetworkopen.2026.0043?utm_source=openai))

That limitation matters. Consensus can clarify professional judgment. It cannot erase uncertainty.

Still, an annual review is a reasonable decision standard because it asks for something modest but frequently missing: a current explanation of why a medication remains part of the plan.

Past benefit is evidence, not permanent proof

One of the most common assumptions in long-term prescribing is that benefit at the beginning establishes benefit indefinitely.

It may. It may not.

A medication’s original purpose should remain part of the record: what symptoms existed before it was started, how severe they were, what changed after starting, what happened after dose changes, and what else changed in the person’s life. Without that history, later decisions are forced to rely on memory, which is often incomplete and easily shaped by fear, hope, or habit.

The same applies to costs. A side effect that seemed tolerable in the first few months may carry a different weight after years of exposure. Sexual effects, emotional flattening, sleep disruption, weight or metabolic changes, cognitive complaints, falls, movement symptoms, sedation, and drug interactions are not interchangeable. Nor are they equally relevant to every medication or every person.

The question is not whether a cost is objectively large enough to count. It is whether the benefit and cost still make sense together.

Why symptoms after a change are hard to interpret

The decision becomes more complicated when a medication is reduced or stopped and familiar symptoms return.

Depression, anxiety, insomnia, agitation, and panic may represent relapse or recurrence. They may also occur as withdrawal symptoms, rebound effects, new problems, medication effects from another drug, or responses to changing life circumstances. More than one explanation can be true at once.

Physical dependence adds another layer. Dependence means the body has adapted to the presence of a medication. It is not the same thing as addiction, which involves compulsive use despite harm. Physical dependence does not prove that a medication was inappropriate. It also does not prove that a person requires it forever. ([rcpsych.ac.uk](https://www.rcpsych.ac.uk/mental-health/treatments-and-wellbeing/stopping-antidepressants?utm_source=openai))

Recent research reinforces why simplistic labels are risky. In a small prospective cohort of 32 adults tapering antidepressants, changes in depression and anxiety scores closely tracked changes in reported withdrawal symptoms over 26 weeks. That finding does not establish that all distress during a medication change is withdrawal, and the study is too small to settle the broader question. It does show why familiar emotional symptoms cannot automatically be counted as proof of relapse. ([pubmed.ncbi.nlm.nih.gov](https://pubmed.ncbi.nlm.nih.gov/42594039/?utm_source=openai))

A useful medication review asks better questions

A serious review is more than asking whether someone wants another refill. It makes the assumptions visible.

  • What problem was this medication originally intended to address?
  • What evidence suggests it is helping now?
  • What costs or tradeoffs may have become easier to overlook?
  • Has the diagnosis, life context, health status, or medication list changed?
  • Could one medication be masking an adverse effect, withdrawal effect, or interaction involving another?
  • What has happened during prior dose changes, missed doses, additions, or discontinuations?
  • What would count as a meaningful improvement, a tolerable tradeoff, or an unacceptable risk?

These questions do not predetermine the answer. They improve the quality of the answer.

A medication review should not be a ritualized path toward stopping. It should not be a ritualized defense of continuing, either.

It is a chance to replace inertia with a current account of benefit, burden, uncertainty, and priorities. That is a more honest foundation for any medication decision.

Written for Dr. Teralyn Sell, PhD
Psychology · Brain Health · Human Behavior

Sources worth knowing

Submit a Comment