Insights: How Therapy Can Go Wrong And What Clinicians Can Do To Make It Go Well – Dr Elizabeth Li

Welcome to Psychology Tools Insights

Psychology Tools Insights explores thought-provoking new research, translating findings into accessible clinical takeaways for mental health professionals. Hear directly from the authors about their latest ideas, and how to effectively incorporate findings into your work, keeping you up-to-date and evolving your practice.

Unfortunately, psychotherapy isn’t always helpful for those that need it. Patients can be left feeling let down, despondent, and even resigned to never seek professional help again. Now a new study has shed light on where psychotherapy is most likely to go wrong – as well as which factors help make it a success.

A team led by Dr Elizabeth Li from University College London (UCL) tracked a diverse mix of 148 adults aged 18-63, who had received psychological treatment from the UK’s National Health Service (NHS). 92 of the individuals said their psychotherapy was helpful, but 37 said it was unhelpful and 19 expressed mixed feelings about their experience. The data was collected online and anonymously, and the findings have been published in the journal Psychotherapy Research.

According to Dr Li patients were “surprisingly upfront with their feelings and experiences”. Key themes that emerged in the study include the patients’ preference for personalized treatment, the importance of a therapeutic alliance, and the demand for depth in therapy.

Preference for personalized treatment

Participants who were dissatisfied with their therapy particularly criticized it for being overly structured and impersonal. For example, one patient complained that “it felt like I was being tailored to the treatment rather than the treatment being tailored to me, that nothing in the session would have changed, no matter who was sitting in my place”.

While some did want structured, action-orientated interventions, such as cognitive behavioral therapy, others called for a deeper investigation and understanding of their unique problems before moving on to goal-oriented solutions. “It doesn't necessarily matter which specific approach or theoretical model is used, or if one approach is more effective than another,” notes Dr Li. "Our research implies that what matters most to patients is possibly having a greater sense of agency and collaborative decision-making in shaping the therapeutic process.”

The team’s findings align with the core idea of control-mastery theory (CMT), which posits that individuals enter therapy with unconscious plans aimed at overcoming their pathogenic beliefs that stem from previous traumatic experiences and achieving adaptive personal goals. According to this theory, patients provide cues – direct or subtle – about what they need in therapy. “For example,” explains Dr Li, “a patient might coach the therapist to respond in a supportive, non-judgmental manner, which contrasts with the criticism they might have experienced in early relationships.

"When therapists respond in ways that align with the patient’s healing goals, it reinforces the patient’s motivation to let go of limiting beliefs, which facilitates the patient’s personal growth and psychological relief. I believe the concepts of patient coaching and therapist responsiveness, though rooted in psychodynamic approaches, can and should be applied across various treatment models, including CBT."

The personalized approach also extends to the mode of delivery: some of the participants found online or telephone therapy helpful. The flexibility of these methods were a major benefit, and some participants felt they could disclose more than in a face-to-face session, while others found it “impersonal.”

Demand for depth

Patients who were dissatisfied or had mixed feelings about their therapy felt that it had merely scratched the surface, and failed to ensure lasting benefits. One patient reported: “I tried to find out what the fundamental problem I have is – the problem ‘roots’… In the end, [it felt like] we were trying to look at other problems (‘branches’), that seem to be a by-product of my main problems.”

Due to the nature of the outcome measures used in public health systems (such as self-report questionnaires), clinicians often prioritize symptom alleviation. However, patients may not always feel ready to finish therapy, even when their symptoms have improved.

To address this problem, Dr Li stresses the importance of clear communication with clients about the goals and scope of the therapy, as well as frequent reviews of their progress throughout treatment: “Lack of depth can be a common issue in time-limited therapies, especially in public health systems like the NHS, where resources are stretched. Due to the busy environment and high workload, many NHS therapies may be required to focus on symptom relief and short-term outcomes, which can mean less time to explore deeper, underlying issues.

“While it’s important to communicate the level of depth achievable within the specific therapeutic approach at the outset, it’s also vital for patients to do their own reflective work outside of therapy. Similarly, if deeper work is required, it may need to be pursued in longer-term or private therapy. However, therapists can still work to create as much depth as possible within the given constraints by being flexible, responsive, and open to exploring emotional complexity when it arises.”

The importance of the therapeutic alliance

Patients’ perceptions of the therapeutic relationship considerably influenced their overall experience of therapy, the researchers found. Therapists who related to them as “unique individuals” and used a “non-judgmental approach and genuine curiosity” were appreciated, while a perceived lack of empathy could significantly affect their engagement with treatment. For example, one patient said her therapist “should have listened more deeply to my words and been more compassionate and patient, especially when I was visibly distressed by what she was saying.”

Patients suggested that the therapeutic bond might be improved if therapists occasionally share their own experiences. Dr Li recalls an occasion when this proved useful: “In one of my psychotherapy research studies, I listened to sessions involving depressed adolescents in short-term psychodynamic psychotherapy. Normally, psychodynamic therapy focuses heavily on exploring emotional experiences and unconscious processes, rather than offering direct advice. Despite this, one adolescent specifically asked the therapist, 'If you were in my situation, what would you do?' – seeking direct advice with an open mind.”

In this case, a CBT approach might feel more natural, as it often involves providing practical suggestions for handling specific situations (such as interpersonal problems). When the psychotherapist took an active stance by offering real-life advice and personal insights, the patient was left “visibly happy” and showed clear improvement as a result of receiving the guidance she sought.

However, care needs to be taken when sharing this type of information. Dr Li warns: "Therapists should carefully observe patient cues to determine when sharing personal insights is appropriate. While some patients may actively seek and benefit from a human-to-human connection, others may see personal sharing as irrelevant, mistimed, self-serving, or patronizing. The key is to recognize when and how to share personal experiences in a way that benefits the patient. The therapist needs to assess whether the patient is seeking empathy, active listening, or validation rather than personal insights or advice. Sometimes, attuned listening and being responsive to the patient’s needs in the moment can be more helpful than sharing personal insights.”

For some clients, even well-timed and relevant personal insights from the therapist may not always be received as intended. Dr Li describes the case of one of her own long-term patients, who has experienced intrusive parenting and a lack of genuine care, and who almost constantly craves validation. When Dr Li has shared relevant personal insights to subtly validate her, the patient responded with a sense of urgency, finding it difficult to focus on anything other than her own emotions or experience. Ultimately, the value of therapists sharing personal experiences will depend on the client’s emotional needs and focus.

Confidentiality conflicts

Another difficulty that can arise in therapy is maintaining the therapeutic bond when confidentiality needs to be broken for safety reasons.

One of the study respondents  described how he didn't want to engage in therapy any more after the therapist called his parents regarding his suicidal urges. Dr Li responds: “This is indeed a challenging situation, especially in an NHS setting where I believe standardized safeguarding procedures must be followed. This is especially true for patients who entered the system before the age of 18, in services such as CAMHS (Child and Adolescent Mental Health Services).”

She adds that the therapist in this situation likely followed guidelines that prioritize patient safety, but they could have handled the conversation with the patient more sensitively, explaining the need for safeguarding in a way that maintains trust. Setting clear expectations about confidentiality and when it might be broken due to safety concerns early on can help manage such situations more effectively. One of Dr Li’s colleagues had a similar experience after meeting with an adolescent's parent to discuss a specific event: “Later, when he met with the adolescent again, he mentioned that he had spoken with their parent about that event. The adolescent responded, ‘I was wondering if you'd be honest with me about discussing this with my mom. If you hadn’t told me today, I would have stopped coming.’”

Therapists need to strike a balance between safeguarding and maintaining trust in the therapeutic relationship. Beyond the issue of confidentiality, this also highlights how patients sometimes ‘test’ their therapists as a way of building trust in them.

When patients feel they can’t be helped

Some of the patients in the study believed the biological nature of their issues meant that therapy would not be able to help them. These patients often emphasized their reliance on self-management and individual resilience, suggesting a preference for addressing issues independently rather than seeking external support. In the words of one respondent: “I think it was good to talk to someone about my problems, but ultimately, I don't think there's anything I can do about the way my brain works.” Patients may abandon therapy if they feel this way – especially if their beliefs have developed in response to unhelpful therapy.

Even when therapists have done everything right, there may be some situations where it is hard for them to make a difference. "There is a group of people who may fall under the concept of the 'p-factor,' a general psychopathology factor,” explains Dr Li. “Some disorders may not respond well to psychotherapy. This is particularly true of schizophrenia, which is largely driven by genetic factors. We’ve also heard terms like ‘difficult’ patients (though this might not be the best way to describe them), such as individuals with complex PTSD, or those who have developed epistemic mistrust [a deep distrust in social communication]. These groups may find it hard to benefit from therapy without significant long-term effort.”

While researchers continue to explore which theoretical orientations work best for particular disorders, Dr Li recommends focusing on communication and flexibility: “The key point is listening to patients and adjusting the approach to meet their needs – rather than sticking rigidly to a specific theoretical model or method– with the therapist being responsive to patient coaching.”

The study was carried out within the San Francisco Psychotherapy Research Group (SFPRG). The full research team: Elizabeth Li (UCL), David Kealy (University of British Columbia), Katie Aafjes-van Doorn (previously Yeshiva University New York, now NYU Shanghai), James McCollum (SFPRG), John T. Curtis (University of California San Francisco & SFPRG), Xiaochen Luo (Santa Clara University), and George Silberschatz (University of California San Francisco & SFPRG).

Read the full paper here