What if therapy is not helping?

Effective treatment does not mean everyone improves

Psychological therapies are recommended because, across well-designed trials, they help more people than control conditions. That average can hide a less comfortable truth: a substantial number of people do not show a clear response to the first treatment they receive. A 2024 meta-analysis brought together 441 trials involving 33,881 participants across eight mental health conditions. Depending on the condition, the estimated response rate to psychotherapy ranged from about 24% to 42%.

Those figures should not be read as an individual forecast, and they do not mean everyone outside the response group was harmed. The trials used a fairly demanding definition of response, usually at least a 50% reduction in symptoms, and differed in treatment, setting and follow-up. What they do show is that non-response is not rare or evidence that a person lacks motivation. Sometimes a good treatment has not yet had enough time; sometimes the target, method or formulation needs to change.

Feeling stuck can mean several different things

Therapy may appear unchanged while something important is shifting underneath. A person might understand their pattern earlier, recover from an episode more quickly or take one difficult action that a symptom questionnaire does not capture. The reverse can also happen: sessions feel thoughtful and emotionally meaningful, but daily avoidance, relationships, sleep or work functioning remain much the same. That is why progress needs more than a general question about whether therapy feels useful.

A review might find that the original goal was too broad, a new stressor has overwhelmed the gains, the treatment is targeting symptoms while the main maintaining pattern sits elsewhere, or the person and psychologist are using different definitions of improvement. Medication changes, physical health, substance use, safety, practical instability and the frequency of appointments can also affect progress. These possibilities call for curiosity and a more precise formulation, not a quick verdict about who is at fault.

The relationship can go off track quietly

Sometimes the difficulty sits in the working relationship. Psychotherapy researchers call these moments alliance ruptures. A rupture can look confrontational, such as openly disagreeing with the psychologist, but it can also be quiet: changing the subject, becoming compliant, saying everything is fine, or gradually deciding that the psychologist does not understand while never quite saying so. For people who expect criticism, rejection or conflict, silence may feel safer than risking an honest response.

A July 2026 meta-analysis found a small association between directly observed alliance ruptures and poorer outcomes. The result is useful but needs restraint: it came from only four publications, covering 301 clients, and an association does not show that a rupture caused the poorer outcome. Ruptures are common and can sometimes become productive when they are recognised and repaired. The clinical point is simpler than the statistics: unspoken disagreement can deprive therapy of information it needs.

Progress measures help, although they are not the treatment

Brief questionnaires can make change, or the lack of it, harder to overlook. They may track symptoms, functioning, quality of life or the therapy relationship across appointments. In a meta-analysis of randomised studies, measurement-feedback systems produced a small average improvement over treatment as usual. The effect was larger among people identified as not on track, which is the group for whom early detection is most likely to alter the course of treatment.

The effect sizes were modest, the studies used different systems and there was substantial variation between results. A score can also miss culture, context and the particular change a person values. Routine measurement is most useful when it supports a real conversation rather than becoming a performance target or another form to complete. It gives the client and psychologist one more source of evidence to place alongside clinical judgement, lived experience and the formulation.

A direct review is part of good therapy

You do not need a polished critique of your psychologist. It is enough to say that you understand more but daily life has not shifted, that the sessions feel repetitive, or that something in the approach is not landing. A useful review returns to a few practical questions: What are we trying to change? What would count as progress? Which parts of therapy have helped, which have not, and what is our current explanation for that pattern? If we continue, what will be different over the next few sessions?

A psychologist should be able to hear that without becoming defensive or asking the client to protect their confidence. The next step might be to sharpen the goals, revisit the diagnosis or formulation, use a different treatment method, change the pace, address avoidance more directly, coordinate with another health professional or recommend someone whose scope better matches the problem. Continuing can be sensible, but it should become an active decision with a clearer rationale rather than the default because both people feel awkward about stopping.

Discomfort is not automatically harm

Therapy can temporarily feel harder when someone approaches a feared situation, discusses painful material, sets a boundary or stops using a familiar coping strategy. That discomfort may be expected and worthwhile, but it should have a clear purpose, be discussed openly and remain connected to consent. A 2025 review of negative effects in psychotherapy emphasised that researchers need to distinguish expected treatment reactions, unwanted side effects, deterioration, incorrectly applied interventions and malpractice rather than placing every difficult experience in one category.

Persistent worsening, increasing shame, pressure to disclose before there is consent, repeated misunderstanding, unclear boundaries or a treatment plan that cannot be explained should not be dismissed with ‘therapy is meant to be hard’. At the same time, changing therapists every time anxiety rises can strengthen avoidance. There is no single rule for when to persist or leave. A careful decision uses progress data where available, the person’s experience, the treatment evidence and an honest discussion about what the current formulation may be missing.

Good therapy does not guarantee quick or steady improvement. It does take a lack of improvement seriously. Being able to pause, name what is not working and revise the plan is not separate from treatment; sometimes it is the piece of treatment that allows the work to move again.