“The patient is not improving” sounds like a finding, but it is still too broad to guide a decision. Which dimension is not changing? For how long? In relation to which goal, measure or expected trajectory? Is functioning stable while symptoms fluctuate, or is there clinically significant deterioration?
When treatment is not progressing as expected, it calls for review rather than an immediate search for someone to blame. The issue may involve the formulation, intervention, dose, engagement, alliance, context or the indicator used to observe change. Often, several factors interact.
First define what “not improving” means
Turn the impression into a clinical question by reviewing at least five dimensions:
- Symptoms: intensity, frequency, duration and variability.
- Functioning: work or study, sleep, relationships, self-care and daily activities.
- Personal goals: what the person wanted to be able to do or face.
- Relevant processes: avoidance, rumination, emotional regulation, skills or other factors in the formulation.
- Experience of treatment: understanding, perceived usefulness, alliance, burden and unwanted effects.
An unchanged score does not always mean that nothing has changed. Functioning may have improved, or new contextual difficulties may be masking progress. Conversely, a positive general impression may miss a specific area of deterioration. Quantitative and qualitative information need to be read together.
The APA statement on evidence-based practice in psychology includes ongoing monitoring of progress and treatment adjustment when needed. Review is therefore not an exception or a failure; it is part of good practice.
Check safety and relevant changes immediately
Before discussing techniques, reassess any risk indicators, rapid deterioration, adverse events, reported medical or medication changes, substance use and major contextual changes. If urgent needs or issues outside the available setting and competence emerge, activate appropriate procedures and consider consultation, collaboration or referral.
Do not wait for the planned review when conditions change significantly. The treatment calendar must not take priority over safety.
A clinical audit in eight questions
1. Is the goal still appropriate and shared?
A goal may have become less relevant, may be too abstract or may have been defined more by the therapist than the person. Ask what change would matter now and how both of you would recognize progress. Shared decision making includes the option to continue, change or not change a plan after discussing alternatives, benefits and limitations.
2. Are the baseline and expected timescale clear?
Without a baseline, normal fluctuation can be confused with lack of progress. Check the starting point, measurement frequency, what would count as clinically meaningful change and whether the observed period is appropriate for the problem and intervention.
Avoid both changing too early and continuing indefinitely. Set a review window before the next cycle of work.
3. Does the formulation still explain the case?
Update priority problems and predisposing, precipitating, perpetuating and protective factors. Ask which new information contradicts the initial hypothesis and which missing fact would change your decision.
Pay particular attention to:
- comorbidities or conditions not previously considered;
- medical, neurocognitive or substance-related factors to investigate in the appropriate setting;
- recent events and persistent stressors;
- resources and protective factors;
- the cultural meaning of the problem and treatment;
- differential diagnosis and alternative hypotheses.
4. Was the intervention delivered as intended?
Before concluding that a strategy does not work, check dose, frequency, continuity, understanding of the rationale and quality of implementation. An intervention that is appropriate on paper may have been fragmented, too complex or introduced without the necessary prerequisites.
This review should not become a way to place responsibility on the patient. It also includes the therapist, the setting, the instructions provided and organizational barriers.
5. Does treatment fit the person and context?
Preferences, values, culture, resources, stage of change and daily burden affect feasibility. The APA guidelines on evidence-based psychological practice encourage adapting the approach to the person’s characteristics and preferences while balancing fidelity to evidence-based methods with clinical responsiveness.
An adaptation needs a rationale. Changing everything to remove difficulty makes treatment incoherent; ignoring the person’s circumstances makes it impractical.
6. What is happening in the alliance?
Unspoken disagreements, divergent goals, tension or poorly shared tasks can slow treatment. Repairing alliance ruptures is associated with better outcomes in the available meta-analytic literature.
Ask directly: “Is there anything about the way we are working that feels unhelpful or that you would like to change?” Feedback is clinically useful only when critical answers can be heard and discussed.
7. Are the data actually being used?
Monitoring is not simply administering measures and filing them away. NHS Talking Therapies service standards describe session-by-session use of standardized measures so that the person and clinician have current information and can identify points at which progress becomes stuck.
A research review of routine outcome monitoring and feedback found small average effects that are more relevant for cases not progressing as expected when feedback is discussed and connected with clinical support tools. Scores do not replace judgment; they make better questions visible.
8. Are there unwanted effects or deterioration?
No improvement and deterioration are not the same. Ask explicitly whether treatment is increasing shame, dependence, conflict, avoidance, stress or other difficulties, and consider alternative explanations. A systematic review of adverse events in psychotherapy trials found that the monitoring and reporting of possible harms remain inconsistent and should become more systematic.
Deterioration requires prompt assessment, not simply continuing until the protocol is expected to work.
From review to decision
After the audit, develop two or three options. For each one, specify the rationale, expected benefits, risks, necessary conditions and review criterion.
Continue
Continuation makes sense when the rationale remains strong, treatment has been delivered adequately, the timescale is still appropriate and no signs of harm are present. Define when and how progress will be reviewed again.
Adapt
Adaptation may mean simplifying a task, changing frequency or delivery, addressing a barrier, repairing the alliance or making the goal more collaborative. The change should remain consistent with the formulation.
Change strategy
A change is appropriate when the data weaken the original rationale, the intervention does not fit or another option has a better balance of expected benefits and risks. Document why the strategy changes and which indicator will be assessed.
Add expertise or services
Consultation, collaboration with other professionals or referral may be needed when different needs, limits of competence or requirements for another level of care emerge. The Italian Psychologists’ Code of Ethics addresses professional responsibility, limits of competence and referral when other specific expertise is required.
End or redefine treatment
If the goals, setting or treatment are no longer appropriate, it may be right to discuss ending or transitioning care. The decision should not be framed as punishment for limited improvement and should include clear information about possible continuity.
A quick review sheet
Complete these fields before discussing the case with the patient or in supervision:
- Agreed goal and chosen indicator.
- Observed course compared with baseline.
- Changes in functioning and context.
- Interventions delivered, dose and response.
- Engagement, preferences and barriers.
- State of the alliance and feedback received.
- New hypotheses or missing information.
- Risks, unwanted effects and protective factors.
- Realistic options with benefits and limitations.
- Shared decision and date of the next review.
If the first three fields cannot be completed, gathering information may be the most useful next intervention.
Biases to keep in view
- Confirmation bias: noticing only information that supports the initial formulation.
- Escalation of commitment: continuing because much has already been invested in treatment.
- Recency bias: giving too much weight to the latest session.
- Attribution bias: explaining the impasse through the person’s motivation while underestimating method and context.
- Loyalty to a preferred technique: fitting the case to the tool rather than the other way around.
- Overconfidence in AI: treating a well-written answer as a reliable assessment.
Supervision, measurement, documentation and open discussion reduce these risks without eliminating them.
Using PsyLab to organize the review
PsyLab can help separate data, hypotheses and decisions, build a de-identified timeline, compare blocking factors and prepare questions for the next session. It does not determine whether treatment is effective and does not replace risk assessment, guidelines, supervision or interdisciplinary collaboration.
A cautious prompt could be:
“Help me organize the review of a treatment that is not progressing as expected. Use only de-identified information: goals, baseline, symptom and functioning course, interventions and dose, engagement, alliance, contextual events and risks already assessed. Separate data from interpretations, identify missing information and compare the options of continuing, adapting, changing or seeking consultation. Do not diagnose or decide for me.”
If the review suggests that engagement is the main issue, explore how to address limited engagement in therapy. If several alternatives emerge, use an explicit structure for choosing the next therapeutic intervention.
Frequently asked questions
After how many sessions can treatment be considered ineffective?
There is no number that applies to every problem and intervention. You need a baseline, expected course, adequate dose, relevant goals and appropriate measures. The review window should be set in advance and brought forward if deterioration or risk emerges.
Does an unchanged score mean that the patient has not improved?
Not necessarily. Consider the measure’s reliability and integrate it with functioning, personal goals, subjective experience and context. At the same time, a positive impression should not override indicators that remain unchanged or worsen.
Should the technique be changed immediately?
First review formulation, goals, delivery, dose, alliance, preferences and barriers. If the rationale remains sound, adaptation or continuation may be appropriate. If the data weaken it, compare alternatives and set a new review criterion.
When is supervision or referral needed?
When the impasse persists, risk or needs outside the therapist’s competence emerge, the therapist’s reactions interfere with understanding the case, or other professional expertise is required. Referral should be explained in terms of the person’s needs, not framed as failure.
Examine possible blocking factors with PsyLab
Organize progress, interventions, alliance and context; compare hypotheses and prepare a collaborative review. Enter de-identified information only and keep clinical decisions under your professional responsibility.