A medication regimen can become more complicated for a perfectly understandable reason.
A person starts a medication. It helps one problem, but a new difficulty appears. Sleep changes. Weight changes. Sexual function changes. Restlessness, nausea, fatigue, tremor, or emotional flattening enters the picture. Another medication is added to address it.
Sometimes that second medication is useful. Sometimes it is the least disruptive option in a difficult situation. And sometimes it becomes the beginning of a prescribing cascade: one medication added to manage an effect, uncertainty, or consequence associated with another.
That sounds straightforward. It isn’t.
A February 25, 2026 consensus statement from the American Society of Clinical Psychopharmacology put this issue into plain view. The 45-member panel agreed that a medication added as an “antidote” for an adverse effect, including a medication added for antidepressant-related sexual dysfunction, should not simply remain in place if it has not meaningfully done the job it was added to do. The larger point matters more than the example: psychiatric regimens need to be read as systems, not as a collection of unrelated prescriptions. ([jamanetwork.com](https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2845497))
The important question is often chronological
When people look at a medication list, they often see the present tense: what is being taken now, at what dose, for what stated diagnosis.
But the more revealing document is often a timeline.
What was present before the first medication? What changed after it began? What shifted after a dose increase? When did the next symptom appear? Was the second medication intended to treat an underlying problem, a side effect, a withdrawal symptom, a rebound effect, or a symptom whose cause was never clear?
Those are not interchangeable categories.
Insomnia, for example, may be part of an underlying condition. It may also emerge after a medication is started, during long-term exposure, after a dose change, or during withdrawal. A medication prescribed for sleep may be appropriate in more than one of those situations. But the meaning of that prescription changes depending on which situation it is addressing.
Without the sequence, it is easy to mistake a treatment consequence for a new disorder, or to overlook that an additional medication was meant to be temporary.
Relief is real. So is the tradeoff.
There is a simplistic story that says every additional psychiatric medication represents a mistake. That is not serious thinking.
Some people have complex conditions, partial responses, multiple symptoms, or adverse effects that require careful management. A second medication can improve quality of life, preserve function, or make a previously helpful medication tolerable enough to continue. The fact that a medication was added does not tell us whether it was inappropriate.
But neither does symptom relief settle the question.
Every addition changes the decision environment. There may be new side effects, interactions, costs, monitoring requirements, practical burdens, and new uncertainty about what is causing what. Over time, a regimen can become difficult to interpret precisely because several changes happened in close succession.
I keep coming back to one question: What problem is this medication solving today, and what problem might it be creating or concealing?
That question applies equally to the first medication and the fifth.
Do not confuse a consensus statement with settled evidence
The 2026 statement is useful because it names problems clinicians and patients encounter all the time: redundant medications, conflicting effects, adverse-effect management, interactions, and regimens that no longer fit their original purpose.
It is also important to understand what it is. It is a Delphi consensus process and selective literature review, not a set of trials proving one universal approach produces better outcomes. The authors explicitly note that empirical trials are still needed to test deprescribing protocols and their outcomes. ([jamanetwork.com](https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2845497))
That distinction is not a technicality. It is the difference between saying, “Experts agree this deserves structured attention,” and saying, “We know exactly what should happen next in every case.” We do not.
The evidence base is especially limited when a regimen is changed after long-term use. A Cochrane review of antidepressant discontinuation studies found that many trials used abrupt stopping or tapers of only a few weeks, often did not assess withdrawal well, and generally did not distinguish withdrawal symptoms from relapse. That leaves important questions unresolved. ([cochrane.org](https://www.cochrane.org/evidence/CD013495_stopping-long-term-antidepressants-people-depression-or-anxiety))
So if distressing symptoms occur after a medication change, it is not intellectually honest to automatically label them relapse. It is also not intellectually honest to automatically label them withdrawal. The timing, symptom pattern, accompanying changes, prior history, and course over time all matter. Sometimes more than one explanation is plausible.
A better way to read a complex regimen
Instead of asking whether a medication list is “good” or “bad,” it helps to ask a series of more precise questions:
- What was each medication originally intended to address?
- What evidence suggests it has helped that target?
- What changed after it was started or adjusted?
- Was another medication added in response to that change?
- Is the added medication still producing a meaningful benefit?
- Are two medications now treating each other’s effects more than they are treating the original problem?
- What would be difficult to interpret if multiple changes occurred at once?
- What outcomes matter beyond symptom scores, including sleep, work, relationships, sexual function, metabolic health, cognition, and daily functioning?
Notice what this framework does not do. It does not begin with a conclusion that more medication is wrong. It does not assume that fewer medications is automatically safer. And it does not turn every adverse effect into proof that treatment failed.
It asks whether the current arrangement is still coherent.
The downstream cost of not revisiting the story
Medication lists have momentum. A prescription can outlast the symptom it was meant to treat. An adjunct can become permanent because nobody revisits its original purpose. A side effect can become folded into a diagnosis. And a person can be left trying to manage the cumulative burden of medications without a shared account of how the regimen got there.
The best medication decisions are not necessarily the newest, the most aggressive, or the most minimalist. They are the decisions that remain legible over time.
If one medication is treating a problem created by another, that may be the right tradeoff. But it deserves to be recognized as a tradeoff, not mistaken for a permanent fact of the person.
Written for Dr. Teralyn Sell, PhD
Psychology · Brain Health · Human Behavior