Signs Your Mental Health Treatment Isn't Working, and What to Ask For

Treatment may not be working if, after a fair trial, your symptom scores have barely moved, you are better but still clearly unwell, or side effects are costing you more than the treatment gives back. A fair trial of an antidepressant is usually four to eight weeks at an adequate dose, and NICE suggests reviewing the plan if there has been no response at all after four weeks of medication, or four to six weeks of therapy.

Some changes should not wait for the next appointment: worsening thoughts of suicide, new agitation or restlessness after starting a medication, or signs of unusually high energy and little need for sleep. Contact your clinician promptly if any of these appear.

How long to give a treatment

Early change is a useful signal. Sleep, appetite, energy and concentration often improve before mood does on an antidepressant, so small shifts in those areas count. A complete absence of change by the review point is different from slow but steady progress.

Measure it, don't guess

How you feel on the day of an appointment is a poor guide to how the last month went. Short, validated questionnaires make progress visible, and NICE suggests considering a validated measure to inform and evaluate treatment, and routine outcome monitoring once treatment starts. Tracking scores over time and adjusting the plan when they stall is often called measurement-based care.

• PHQ-9 for depression: nine questions, scored so that 5, 10, 15 and 20 mark the thresholds for mild, moderate, moderately severe and severe depression.
• GAD-7 for anxiety: seven questions, validated for screening generalized anxiety disorder and rating its severity.

Fill one in before treatment starts and every few weeks after, at roughly the same point in the week so the scores are comparable. If your clinician does not use them, you can still bring your own scores to visits.

How big a change counts

Scores bounce around from week to week, so a one- or two-point shift may be noise. In a study that tracked PHQ-9 scores during depression treatment, the smallest clinically important change for an individual was estimated at 5 points on the 0 to 27 scale. That study enrolled older adults with late-life depression, so treat the number as a rough guide rather than a rule, but it is a useful reference: a drop of five or more points is more likely to be real than a drop of two. PHQ-9 and GAD-7 scores explained covers how to read both scales.

A simple tracking template

A table like this, kept on your phone or on paper, gives your clinician a month of information in a minute.

Add one line about functioning each time, such as "made it to work four days" or "saw a friend". Changes in what you can do often matter as much as the score.

Reading the pattern, not one number

A month or two of scores usually falls into one of a few patterns. Each suggests a different question for your clinician.

None of these patterns is a diagnosis. They are a way to turn a vague sense that things are not right into a specific question your clinician can act on.

Partial response is not the finish line

Research usually defines "response" as at least a 50% drop in symptom scores. "Remission" means symptoms have fallen to a minimal level. The difference is not academic: in the STAR*D study, people who entered follow-up in remission had lower relapse rates than those who entered with residual symptoms.

So if you feel "better, but not myself," that is worth saying plainly. Residual low mood, poor sleep, low motivation or persistent worry are reasons to adjust the plan, not proof that this is as good as it gets.

A good review ends with a concrete plan: what will change, what you should notice and by when, and the date of the next check. If a visit ends with "let's give it more time" and no date or measure attached, ask what more time means and what would prompt a change. Waiting can be the right decision, but it should be a decision rather than a default. What a psychiatrist considers when antidepressants haven't worked covers the options from dose changes to add-on treatments.

Signs the plan needs to change

• Your scores are flat after a fair trial, or rising.
• You are functioning a little better but still missing work, avoiding people, or struggling with basic tasks.
• Improvement came and went within a few weeks.
• You keep skipping doses or sessions, which is often a sign the plan does not fit your life or your side effects.
• Therapy sessions feel unfocused, with no shared goals or homework, and no review of progress.
• Nobody has asked about side effects, sleep, alcohol use or new stressors in a while.

Any one of these is a reason for a review, not necessarily a new treatment. A review may find a dose that was never raised, an untreated sleep problem, or a diagnosis that needs a second look.

When therapy is the main treatment

Therapy can stall for different reasons than medication, and the signs are different too. These are worth noticing:

• You and your therapist have never agreed on specific goals, or you could not say what they are.
• Sessions are mostly catching up on the week, with no skills, practice or homework between them.
• Progress has never been reviewed, with or without a questionnaire.
• You hold back important things because you do not feel safe or understood.
• The approach does not fit the problem, for example general talk therapy for OCD, where a specific method is the standard.

The relationship itself matters. A meta-analysis of 295 studies covering more than 30,000 patients found a consistent positive association between the strength of the therapeutic alliance and treatment outcome across approaches. That is an association rather than proof that the relationship causes improvement, but it supports raising the issue if you do not feel understood. Measuring progress also helps a little: a meta-analysis of 58 studies found that giving therapists regular progress feedback had a small effect on symptom reduction, including for patients who were not on track, and a small favorable effect on dropout. The effect was modest, so feedback is a support, not a fix.

How many sessions a reasonable course takes depends on the problem; how long therapy takes gives typical ranges, and CBT vs psychodynamic therapy explains why the method matters for some conditions.

Timelines differ by condition

The four-to-eight-week rule of thumb for antidepressants does not apply to every condition or every medicine.

• ADHD medication: stimulant doses are adjusted much faster. The Vyvanse label, for example, allows dose adjustments at approximately weekly intervals, so a stimulant trial can often be judged within weeks, using school or work feedback and rating scales.
• OCD: both the expected timeline and the right kind of therapy are specific to OCD, so ask your clinician what a fair trial looks like for you. See OCD treatment options compared and what to expect from ERP.
• Bipolar disorder: progress is judged over longer periods, because the goal includes preventing future episodes, not just lifting the current one. Bipolar disorder medications compared covers the options.
• Anxiety treated with buspirone or an SSRI: see the timing table above; taking these only as needed does not give them a fair trial.

If you are not sure which timeline applies to you, ask your clinician directly: "By when should I expect to notice a difference, and what would we do if I don't?"

Is it the treatment, or something else?

When progress stalls, the treatment itself is only one possible reason. NICE asks clinicians to look first at personal, social and environmental factors, physical and other mental health conditions, and problems with sticking to the plan, such as missed doses or sessions. If response is still poor after those are addressed, the guideline says to review the diagnosis itself.

In practice, that review often turns up something fixable:

• A dose that was started low and never increased.
• Insomnia, which can keep depression and anxiety going on its own.
• Regular alcohol or cannabis use, which can disturb sleep and mood and make it hard to judge the treatment.
• A medical condition or another medication contributing to symptoms.
• Periods of high energy or racing thoughts that point toward bipolar disorder, which needs a different approach; antidepressant labels ask prescribers to screen for it.
• A therapy approach that does not match the problem, or a therapist you do not trust enough to be candid with.

The site's article on signs you may need a new psychiatric medication covers medication-specific signals, and medication, therapy or both explains when adding the other approach tends to help.

Is the level of care right?

Sometimes the treatment is reasonable but the setting is not intensive enough. Weekly therapy and monthly medication visits may not be enough if you are missing most days of work or school, cannot keep yourself safe, or have needed emergency care recently. Options between outpatient visits and hospitalization exist, such as intensive outpatient and partial hospitalization programs. Levels of psychiatric care explained describes each one and who it suits.

The reverse also happens. If you are stable and functioning well, a step down to less frequent visits is a normal part of good care, not a sign that your clinician is losing interest.

Is it the fit with your clinician?

Not every stalled treatment is a clinical problem. Sometimes the clinician is competent but not a good match: they rush visits, do not follow up on what you raised last time, or make you feel judged. Other times there are real warning signs, such as prescribing without discussing risks, refusing to coordinate with your other clinicians, or dismissing side effects. Psychiatrist red flags lists the ones that justify moving on.

Before you switch, it is usually worth one direct conversation, using the script below. Many problems are fixable once named. If they are not, changing clinicians is a reasonable decision, and you do not need anyone's permission to make it.

Side effects you should not just put up with

Some early side effects fade within weeks, and starting low and increasing slowly helps many people. Others are worth raising promptly, because there are usually ways to manage them:

• Sexual side effects, which are common with SSRIs and often go unreported; labels ask prescribers to ask about them specifically.
• Emotional numbness or feeling "flat."
• Persistent insomnia, daytime sedation, or weight change that affects your health or confidence.
• Any side effect that is making you consider stopping.

The site's SSRI side-effect toolkit and guide to managing side effects with your prescriber explain the usual options. Do not stop a medication suddenly to escape side effects; labels recommend gradual reduction, and stopping abruptly can cause discontinuation symptoms. Talk to your prescriber first.

Warning signs that need urgent contact

Antidepressant labels list symptoms that can appear when treatment starts or the dose changes: anxiety, agitation, panic attacks, insomnia, irritability, hostility, aggressiveness, impulsivity, akathisia (an intense inner restlessness), hypomania and mania. The FDA boxed warning notes an increased risk of suicidal thoughts and behaviors in children, adolescents and young adults in short-term trials, and advises close monitoring for anyone starting treatment.

Contact your prescriber the same day, or seek urgent care, if you notice:

• New or worsening thoughts of suicide or self-harm.
• Severe restlessness, agitation or panic that is new since a medication change.
• Racing thoughts, much less need for sleep, unusual risk-taking or grandiosity, which can signal mania or hypomania.
• Agitation, muscle twitching, high temperature or confusion after combining serotonin-acting drugs or supplements, which can signal serotonin syndrome.

If you are thinking about suicide or are in crisis, call or text 988 (the 988 Suicide & Crisis Lifeline) or go to the nearest emergency room.

Teens, young adults and families

For younger patients, the timing of these checks matters more. In the pooled trials behind the antidepressant boxed warning, the increased risk of suicidal thoughts and behaviors was seen in patients age 24 and younger, and NICE suggests a first review one week after starting an antidepressant for people aged 18 to 25. Parents of teens, and partners or roommates of young adults, are often the first to notice agitation, withdrawal or a change in sleep. Agree in advance who will contact the prescriber if they see one of the warning signs above. For teens, the parent's guide to teen depression treatment covers what monitoring should look like.

What to bring to a review visit

A review goes better when the facts are in front of you both. Before the visit, gather:

• Your questionnaire scores since the start of treatment, or the tracking table above.
• The current dose and the date of the last change, plus any doses you missed.
• Side effects, when they began, and whether they are easing.
• Changes in sleep, alcohol or other substance use, and any new medical problems or medications.
• Two or three concrete examples of what is still hard, such as tasks you are avoiding or days you could not work.
• Your own view of what you want to change, even if it is only "I want to feel more like myself."

If you see more than one clinician, such as a prescriber and a therapist, it helps to tell each of them what the other has suggested. Coordinated care works best when everyone is looking at the same picture.

How to raise it with your clinician

Clinicians want to know when treatment is not working; it changes what they do. A short, specific message is easiest to act on. You might say:

"I've been on this dose for six weeks. My PHQ-9 has dropped only three points since we started. I'm sleeping better, but I still can't concentrate at work and I'm avoiding friends. I'd like to talk about whether to adjust the dose, change the medication, or add therapy. What would you suggest, and how will we know if it's working?"

Useful questions to add:

1. Was this an adequate trial, in dose and in time?
2. Could something else be getting in the way, such as sleep, alcohol, a medical problem, or the diagnosis itself?
3. What is the next step, and when will we review it?
4. Would adding or changing therapy help?

If your clinician disagrees

Sometimes the answer is that the plan is working and needs more time. That can be right, but it should come with reasons and a date. Ask what they are seeing that you are not, what specifically would change their mind, and when you will review it again. Write the answer down. If you still disagree at that point, a second opinion is a normal next step, not a betrayal of your current clinician.

If the answers are vague, or concerns are repeatedly brushed aside, a second opinion on your psychiatric medication is reasonable. If you decide to change clinicians, switching psychiatrists without a gap in prescriptions explains how to avoid running out of medication.

At this practice, care starts with a free discovery call that covers care options, scheduling and insurance, and individual clinical care begins with the Patient Interest Form. Ongoing care can include psychiatric evaluation and medication management alongside psychotherapy. The Choosing Psychiatric Care guide covers other starting points.

Frequently asked questions

Should I stop my medication if it is not working?

No. Talk to your prescriber first. Stopping suddenly can cause discontinuation symptoms, and the next step may be a dose change, a switch or an add-on that is best planned together.

Is it normal to feel worse before better on an antidepressant?

Some people have early side effects such as jitteriness or upset stomach, which often ease. New agitation, severe restlessness or worsening suicidal thoughts are different and need prompt contact with your prescriber.

How do I know if therapy is working?

You should have clear goals, a sense of what you are practicing between sessions, and measurable change over time. NICE suggests discussing the plan if there is no response after four to six weeks.

What is the difference between response and remission?

Response usually means symptoms have dropped by at least half. Remission means symptoms have fallen to a minimal level. Remission is the goal.

Will my clinician be offended if I say treatment is not working?

A good clinician wants to know, because it changes the plan. Bring specific examples or questionnaire scores to make the conversation easier.

How many points does my PHQ-9 need to drop to count as real improvement?

There is no single cut-off, but one study of older adults in depression treatment estimated the smallest clinically important change for an individual at about 5 points. Many clinicians also look at whether you have reached the mild range or below, and at how you are functioning, not just the score.

My therapist is kind but I am not getting better. Should I switch?

Raise it first. Ask to agree on specific goals, review progress together, and discuss whether a different method suits your problem better. If nothing changes after that conversation and a few more sessions, switching is reasonable. How long therapy takes can help you judge timing.

Sources

1. Mental Health Medications. National Institute of Mental Health, 2023. nimh.nih.gov
2. Depression in adults: treatment and management (NG222), Recommendations. National Institute for Health and Care Excellence, 2022. nice.org.uk
3. Wellbutrin XL (bupropion extended-release) prescribing information. DailyMed / U.S. FDA label, 2026. dailymed.nlm.nih.gov
4. Generalised anxiety disorder and panic disorder in adults: management (CG113), Recommendations. National Institute for Health and Care Excellence, 2011. nice.org.uk
5. The PHQ-9: validity of a brief depression severity measure. Journal of General Internal Medicine (Kroenke et al.), 2001. pubmed.ncbi.nlm.nih.gov
6. A brief measure for assessing generalized anxiety disorder: the GAD-7. Archives of Internal Medicine (Spitzer et al.), 2006. pubmed.ncbi.nlm.nih.gov
7. A network meta-analysis of the effects of psychotherapies, pharmacotherapies and their combination in the treatment of adult depression. World Psychiatry (Cuijpers et al.), 2020. pubmed.ncbi.nlm.nih.gov
8. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report. American Journal of Psychiatry (Rush et al.), 2006. pubmed.ncbi.nlm.nih.gov
9. Lexapro (escitalopram) prescribing information. DailyMed / U.S. FDA label, 2023. dailymed.nlm.nih.gov
10. Monitoring depression treatment outcomes with the patient health questionnaire-9. Medical Care (Löwe et al.), 2004. pubmed.ncbi.nlm.nih.gov
11. The alliance in adult psychotherapy: A meta-analytic synthesis. Psychotherapy (Flückiger et al.), 2018. pubmed.ncbi.nlm.nih.gov
12. Using progress feedback to improve outcomes and reduce drop-out, treatment duration, and deterioration: A multilevel meta-analysis. Clinical Psychology Review (de Jong et al.), 2021. pubmed.ncbi.nlm.nih.gov
13. Vyvanse (lisdexamfetamine dimesylate) prescribing information. Takeda Pharmaceuticals / DailyMed (NLM), 2026. dailymed.nlm.nih.gov

Educational information, not medical, tax or legal advice. Decisions about diagnosis, medication and treatment belong with a clinician who knows your history.