Therapy that is working does not always feel good. Trauma processing, exposure work, and honest examination of long-held patterns all produce weeks that feel worse than the weeks before them. This makes the question harder than it looks: how do you distinguish productive difficulty from a treatment that is simply not going anywhere?
The distinction is worth making carefully, because both errors are costly. Leaving too early means abandoning something that was about to work. Staying too long means months or years spent on an approach that was never matched to the problem.
Routine outcome monitoring — brief symptom measures administered each session, with results fed back to the clinician — consistently improves outcomes, and the effect is largest for clients who are not responding. Lambert and colleagues have shown across multiple trials that alerting therapists to off-track cases substantially reduces deterioration rates.
The practical implication is simple: if nobody is measuring, nobody can tell whether it is working. A clinician who tracks symptoms with a standard measure such as the PHQ-9 or GAD-7, and who reviews the trend with you, is far better positioned to catch a stalled course early.
Signs the approach may need adjusting
The plan is hard to name
Several months in, you cannot yet say what method is being used or what the goal is. Sessions consist of updating your therapist on the week. Supportive check-ins have real value in their own right — and if you came in for a specific condition with an established protocol, it is worth asking whether that protocol is also part of the plan.
The same conversation, repeatedly
You describe the same pattern, receive the same reflection, and leave with nothing new to do. Progress in most modalities involves something changing between sessions — a skill practiced, an exposure attempted, a belief tested.
The condition-specific protocol has not come up
Some conditions have well-established first-line treatments — ERP for OCD, a trauma-focused protocol for PTSD, interoceptive exposure for panic. Supportive work has real value alongside these, and it is worth asking whether the condition-specific protocol is also part of the plan, since that is where the strongest outcome evidence sits.
Feeling worse, without a shared explanation
Feeling worse during trauma processing is normal, and it is usually a stage rather than a setback. What helps is having a shared explanation for it — a sense of where you are in the arc and what comes next. If that explanation is not yet in place, asking for it is the natural next step.
The conversation has not changed the plan
You said the work was not helping, and the plan has stayed the same since. Feedback like that is genuinely useful clinical information, and most clinicians will adjust once they have it — so if nothing has shifted after a clear conversation, that is worth weighing.
Signs it is working, slowly
Progress is often visible in function before it is visible in feeling. You are sleeping better, or avoiding one fewer situation, or recovering from a bad day in two days instead of ten. You notice a pattern while it is happening rather than a week later. These are real gains, and they routinely precede any change in how you would rate your mood.
Structured protocols in particular have expected shapes. Prolonged Exposure often produces a temporary rise in distress before it falls. ERP is uncomfortable by design. If your clinician predicted the difficulty in advance and can locate where you are in the arc, the difficulty is probably the treatment rather than a failure of it.
Raise it directly first
Clinicians are trained to work with this conversation, and research on the therapeutic alliance suggests that repairing a rupture often strengthens the work. It is a normal part of treatment, not an insult. Be concrete: name what has not changed, over what period, and what you expected.
How to open the conversation
- "I want to talk about progress. I do not think X has shifted in four months. How do you see it?"
- "What method are we using, and is it the standard approach for this?"
- "Can we set a specific goal and a date to review whether we hit it?"
- "Would you consider tracking symptoms with a standard measure so we can see the trend?"
- "If this is not your primary area, is there someone you would refer me to?"
When to change providers
Consider a change when you have raised the concern clearly and the plan has stayed the same; when your condition has an established first-line treatment that would be better delivered by someone trained in it; when trust has not recovered after an honest conversation; or when you have been in treatment for a year or more without being able to name a concrete gain.
Switching is not failure, and it does not erase the work you have done. What you learned about yourself travels with you. Ask for a referral, ask for a summary of the work to hand to the next clinician, and be specific with the next provider about what has already been tried — "I did eighteen months of insight-oriented work and my compulsions did not change" tells a well-trained clinician a great deal.
- Uncomfortable is not the same as ineffective — a shared explanation is what tells them apart.
- Look for function changing before feeling changes.
- Being able to name the method is a good marker of a well-matched plan.
- Raise it directly and concretely before you leave.
- A condition with a first-line treatment deserves that treatment.
References
- Lambert, M. J., Whipple, J. L., & Kleinstück, M. (2018). Collecting and delivering progress feedback: A meta-analysis of routine outcome monitoring. Psychotherapy, 55(4), 520–537.
- Eubanks, C. F., Muran, J. C., & Safran, J. D. (2018). Alliance rupture repair: A meta-analysis. Psychotherapy, 55(4), 508–519.
- Flückiger, C., Del Re, A. C., Wampold, B. E., & Horvath, A. O. (2018). The alliance in adult psychotherapy: A meta-analytic synthesis. Psychotherapy, 55(4), 316–340.