Blog

Pybis, J., Saxon, D., Hill, A., & Barkham, M. (2017). The comparative effectiveness and efficiency of cognitive behaviour therapy and generic counselling in the treatment of depression: Evidence from the 2nd UK National Audit of psychological therapies. BMC Psychiatry, 17, 215. DOI 10.1186/s12888-017-1370-7

Over a decade ago the United Kingdom (UK) invested large sums of public dollars to fund the Increasing Access to Psychotherapy (IAPT) program. In IAPT, most patients receive cognitive behavioral therapy (CBT) as first-line treatment for depression or anxiety, and may receive generic counseling as second line treatment. One of the admirable aspects of IAPT is that the program consistently assesses outcomes, makes its data available for analyses, and publishes yearly reports on their outcomes. In this very large study, Pybis and colleagues assess whether CBT and generic counseling have different outcomes for patients with depression or anxiety. Over 33,000 patients who received treatment at one of 103 sites were in the study. Most patients (about 23,000) receiving CBT, and the others (about 10,000) receiving generic counseling. Two-thirds of the patients were female, most (84%) were white British, and the mean age was 41 (SD = 13.86). CBT focused on changing negative thoughts and behaviors in order to improve depressive symptoms. Generic counselling was harder to define, though the authors described these therapists as practicing in an integrative manner by bringing skills from training in different forms of psychotherapy. Generic counseling therapists did not focus on giving advice or opinions, but rather on helping clients understand themselves better. Pre- to post-treatment effect sizes for CBT (0.94 [0.92, 0.95]) and generic counseling (0.95 [0.92, 0.98]) were equivalent for depression outcomes. In CBT 50.4% of patients reliably improved, whereas 49.6% reliably improved if they received generic counseling. The average number of sessions attended by patients in the two treatments (CBT = 8.9 [6.34]; counseling = 7.5 [5.54]) were also equivalent. However, there were significant site effects. That is, a moderate and significant amount of patient outcomes (15%) could be accounted for by the site at which they received treatment (i.e., some sites or clinics had better outcomes than others).

Practice Implications

Generic counseling as provided in the IAPT in the UK was as effective as structured CBT for reducing symptoms of depression. However, almost half of patients did not improve in either treatment. Generic counseling was likely a label used to describe integrative psychotherapy that followed principles from a variety of psychotherapies that were based on psychological principles. There were much larger site/clinic effects than treatment modality effects, so that clients in some clinics had better than clients who received treatment in other clinics. This is consistent with research on therapist effects that show that some therapists are more effective than others, regardless of their orientation. This research suggests that training therapists to be more effective by improving their facilitative interpersonal skills may yield better outcomes for clients.

Cameron, S. K., Rodgers, J., & Dagnan, D. (2018). The relationship between the therapeutic alliance and clinical outcomes in cognitive behaviour therapy for adults with depression: A meta‐analytic review. Clinical Psychology and Psychotherapy, DOI: 10.1002/cpp.2180.

The therapeutic alliance refers to the collaborative agreement between therapist and client on the tasks of therapy (homework, treatment approach, intervention style) and goals of therapy (to reduce depressive symptoms, to improve interpersonal relationships, to cope better with stress), plus the emotional bond between therapist and client. The alliance is part of a larger concept of therapeutic relationship that also includes the real relationship between client and therapist and the transference relationship (maladaptive relational patterns in the client based on a history of relationships with parental figures). The alliance is thought to be a common factor across different therapeutic orientations, including cognitive behavioral therapy (CBT), time-limited psychodynamic psychotherapy (TLPP), and interpersonal psychotherapy (IPT). In fact, the alliance is known to have a moderate and robust relationship to client outcomes regardless of who rates the alliance (therapist, client, observer), which measure is used, and when in therapy the alliance is rated (early, middle, late). Although Beck emphasized the alliance as a key therapeutic principle in CBT, some CBT writers argue that the alliance is not so important. In this study, Cameron and colleagues conducted a meta analysis of 13 randomized controlled trials that assessed the relationship between therapeutic alliance and CBT outcomes for depression in adult clients. The overall mean correlation between therapeutic alliance and outcome was r = 0.26 (95% CI [.19–.32]), which indicates a moderate and significant relationship. This is very close to the value found in a larger meta analysis of over 200 alliance – outcome studies.

Practice Implications

The study demonstrates the importance of the therapeutic alliance to client outcomes in CBT. The association was at similar levels to those found in other types of therapy. Therapists conducting CBT should attend to building and maintaining an alliance, which provides a context to facilitate CBT interventions. If a client is not completing homework for example, it is likely that there is a lack of agreement on tasks of therapy, and this part of the alliance may need to be renegotiated. Therapists may also benefit from routinely assessing the alliance in therapy with their clients on a session by session basis using short and easy to use measures. Reviewing these scales regularly can alert a therapist to potential problems in the alliance and the need to repair any tensions or ruptures.

Cologan, J., Schweiter, R.D., & Nolte, T. (2017). Therapist reflective functioning, therapist attachment style, and therapist effectiveness. Administration Policy and Mental Health, DOI: 10.1007/s10488-017-0790-5.

Differences between therapists account for about 8% of patient outcomes, which is a moderate effect and therefore an important factor. Constructs such as therapist personality characteristics and facilitative interpersonal skills may play a key role in how effective therapists can be with their clients. An important therapist quality might be reflective functioning, or mentalization. Reflective functioning refers to the ability to conceptualize, identify, and understand mental states in oneself and in others, and how mental states affect behaviour and functioning.  For example, reflective functioning is the basis for predicting others’ behaviors, understanding social nuances and others’ intentions, and also one’s own behaviors and internal experiences. Fundamentally for a therapist, reflective functioning is necessary for empathy, which is a key therapeutic quality. Another key issue for therapists might be their own attachment security, or their characteristic ways of relating to others in interpersonal relationships. Securely attached therapists (those who have a positive view of self and others in relationships) may be able to develop a better therapeutic alliance with clients. Insecurely attached therapists (those who are avoidant in relationships or who are preoccupied in relationships), may struggle to a greater extent with developing and maintaining an alliance. In this study, Cologan and colleagues assessed reflective functioning and attachment security in 25 therapists from different theoretical orientations who treated 1001 adult clients who mostly had problems with depression or anxiety. Client outcomes were measured pre and post treatment. On average clients experienced a reduction in their symptoms after psychotherapy. Clients of therapists with higher levels of reflective functioning experienced better outcomes. Therapist attachment insecurity did not have a direct effect on client outcomes.

Practice Implications

As with other studies, therapists in this study varied in their outcomes, so that some had better outcomes than others. Level of therapist reflective functioning (ability to mentalize) accounted for a large proportion of this difference. Therapists who had greater skills with understanding their own and clients’ behaviors in terms of mental states (intentions, motivations, psychological and emotional needs, internal conflicts) likely were better able to empathize and develop an alliance with their clients. These are skills that therapists can learn with practice, consultation, personal therapy, and training.

Eubanks, C. F., Burckell, L. A., & Goldfried, M. R. (2017, December 21). Clinical consensus strategies to repair ruptures in the therapeutic alliance. Journal of Psychotherapy Integration. Advance online publication.

Research is clear that the therapeutic alliance (i.e., agreement on tasks and goals of therapy, and the bond between client and therapist) is an important predictor of client outcomes across theoretical orientations. It is also clear that ruptures or strains in the alliance occur often and can have a negative effect on client outcomes. One can define two types of ruptures: (1) withdrawal ruptures, in which the client moves away from the therapist by shutting down, changing the focus, or not completing session assignments; and (2) confrontation ruptures, in which the client moves against the therapist so that the relationship quality is low, the client is not collaborative, and the client does not agree with the goals of therapy. Repairing alliance ruptures can have a positive effect on client outcomes, and therapists can learn to repair alliance ruptures. What are the best strategies that a therapist can use to repair alliance ruptures? In this study of expert consensus, Eubanks and colleagues surveyed clinicians in three broad and different surveys. In the first survey, the authors asked 330 professional social workers and psychologists from a variety of theoretical orientations to describe situations in which they encountered alliance ruptures in clinical practice. The researchers categorized situations described by clinicians as withdrawal ruptures or as confrontation ruptures, and then the authors selected those scenarios that best represented each type of rupture. In a second independent survey, 177 clinicians indicated how they would advise a colleague seeking consultation to respond to each scenario of a therapeutic alliance rupture. Clinicians generated between 35 and 45 strategies to repair each type of alliance rupture. In the final part of the survey, training directors in psychology and social work programs nominated peer experts to rate the strategies for alliance repair, so that 134 peer-nominated expert clinicians provided ratings. There was a high level of consensus among experts such that between 55% and 74% agreed on effective strategies to repair alliance ruptures. Experts agreed that during the session in which the alliance rupture occurred therapists should: explore and empathize with the client`s anger at the therapist, and validate or legitimize the client`s position on the issue related to the rupture. Experts also agreed that in future sessions clinicians can use other strategies like: helping the client manage and cope with painful feelings related to the rupture, helping the client clarify and explore their emotions related to the rupture, and exploring the meaning and patterns of problematic relationships outside of therapy.

Practice Implications

Experts agreed that the best strategies to repair therapeutic alliance ruptures were to deal with the therapeutic bond (e.g., explore and empathize with the client`s anger at the therapist) and to validate the client`s position on the issue related to the rupture. Other strategies like helping the client cope with their reactions and feelings, and exploring the meaning and patterns related to the client`s response were also rated as helpful. Less helpful strategies included therapists communicating about the limits of therapy, and therapist self-disclosure of their reaction to the rupture.

Totura, C.M.W., Fields, S.A., & Kraver, M.S. (2018). The role of the therapeutic relationship in psychopharmacological treatment outcomes: A meta-analytic review. Psychiatric Services, 69, 41-47.

There is evidence to suggest that pharmacological treatments are effective for a wide range of disorders. However, a high level of adherence to taking psychotropic medications is necessary in order for them to have a chance of working. Medical interventions in general do not work well when patients are non-adherent to the regimen, and non-adherence is a significant problem in medicine. Treatment adherence is particularly problematic in those with a mental health condition. Low adherence may have to do with problems with the medications themselves, like unpleasant side effects. And low adherence also may be due to issues related to mental health impairment, like low motivation and problems with reasoning. A particular issue in mental health treatment is the manner in which patients receive the medication. Unlike some medical interventions, psychotropic medications are often taken by patients on their own and away from the clinic or hospital. In psychotherapy, we know that a good therapeutic alliance improves outcomes partly because a good alliance provides a context within which psychological interventions can work (i.e., clients may be more adherent to the treatment recommendations) and partly because the alliance itself may be therapeutic. In this meta analysis, Totura and colleagues examine if there is an association between the therapeutic alliance and mental health outcomes for patients who receive pharmacological interventions for their mental illness symptoms. Eight studies of 59 samples representing over 1,000 patients were included. Four studies were of pharmacological treatment for affective disorders, two for schizophrenia, and two for mixed diagnoses. The results indicated a statistically significant and moderate effect: z = .30 (CI=.20, .39, SE=.048, z=6.192, p=.05), such that greater therapeutic alliance predicted better mental health outcomes among patients receiving pharmacotherapy.

Practice Implications

Higher quality of the physician-patient relationship was related to better mental health treatment outcomes for patients taking pharmacotherapy. The therapeutic alliance appears to be just as import in pharmacological treatment as it is in psychotherapy. It is possible that a good alliance with the provider may increase patient adherence, which may lead to better outcomes. It is also possible, however, that the alliance itself is therapeutic. That is, negotiating an alliance and repairing alliance tensions may lead to positive changes in patients’ ability to cope with emotions and to make the most of their social supports. The results also suggest the importance of training physicians in communication skills to improve therapeutic relationships.

Anderson, T., Crowley, M. E. J., Himawan, L., Holmberg, J. K., & Uhlin, B. D. (2016). Therapist facilitative interpersonal skills and training status: A randomized clinical trial on alliance and outcome. Psychotherapy Research, 26(5), 511-529.

Research on therapist effects indicates that there are differences between therapists so that some therapists are more effective than others. Therapist effects account for about 9% of client outcomes, which represents a moderate and therefore important effect. Differences between therapists do not seem to be accounted for by differing levels of adherence to or competence in delivering a manualized treatment. However, some researchers argue that therapist effects can be accounted for by differing level of facilitative interpersonal skills. That is, therapists vary in the level of interpersonal skills, and this difference accounts for a significant proportion of client outcomes. Therapist facilitative interpersonal skills might include: empathy, positive regard, warmth, ability to establish and repair therapeutic alliances, verbal fluency, emotional expression, and the ability to enhance client expectations of improvement. In this unique analogue study, Anderson and colleagues selected 23 “therapists” who were rated as very high or as very low on facilitative interpersonal skills. For example, highly skilled “therapists” scored high on a self-report measure of social skills and also demonstrated high interpersonal skills in their responses to video vignettes of therapy. Therapists also differed on their training status: half of the “therapists” were advanced clinical psychology graduate students, and the other half were graduate students from other programs (social sciences, humanities) who had no clinical training at all. The 66 clients were volunteers from a large undergraduate student research pool who met diagnostic criteria for a mental disorder (anxiety or depression) and were moderately to highly distressed. Clients were randomly assigned to receive treatment or to a wait-list control condition, so that 46 clients (2 per therapist) received treatment and 22 received no treatment. Compared to those in the control condition, clients who received treatment on average improved in terms of level of distress, regardless of which “therapist” they were assigned to. The training status of “therapists” (those with clinical training versus those without clinical training) had no effect on client outcomes or on the therapeutic alliance. Compared to “therapists” with low facilitative interpersonal skills, those with high interpersonal skills (regardless of training status) had significantly better client outcomes and significantly higher levels of the alliance.

Practice Implications

This was an analogue study in which some “therapists” were non-clinicians, so one must take the results with a grain of salt. Nevertheless, clients started out distressed, had a diagnosable disorder, and on average they achieved significant reduction in distress if they received therapy. Whether “therapists” had any clinical training did not affect outcomes, that is, non-clinical “therapists” did just as well as clinical trainees. However, higher “therapist” facilitative interpersonal skills regardless of training status lead to better client outcomes. These findings provide support for the notion that a therapist who is: empathic, warm, able to establish and repair therapeutic alliances, verbally fluent, emotionally expressive, and able to enhance client expectations of improvement will be more effective in reducing their clients’ levels of distress.

Author email: andersot@ohio.edu