Blog
The Psychotherapy Practice Research Network (PPRNet) blog began in 2013 in response to psychotherapy clinicians, researchers, and educators who expressed interest in receiving regular information about current practice-oriented psychotherapy research. It offers a monthly summary of two or three published psychotherapy research articles. Each summary is authored by Dr. Tasca and highlights practice implications of selected articles. Past blogs are available in the archives. This content is only available in English.
This month...
…I blog about psychotherapy for adult depression, efficacy of CBT for perfectionism, and a measure of ruptures in the therapeutic alliance.
Type of Research
Topics
- ALL Topics (clear)
- Adherance
- Alliance and Therapeutic Relationship
- Anxiety Disorders
- Attachment
- Attendance, Attrition, and Drop-Out
- Client Factors
- Client Preferences
- Cognitive Therapy (CT) and Cognitive-Behavioural Therapy (CBT)
- Combination Therapy
- Common Factors
- Cost-effectiveness
- Depression and Depressive Symptoms
- Efficacy of Treatments
- Empathy
- Feedback and Progress Monitoring
- Group Psychotherapy
- Illness and Medical Comorbidities
- Interpersonal Psychotherapy (IPT)
- Long-term Outcomes
- Medications/Pharmacotherapy
- Miscellaneous
- Neuroscience and Brain
- Outcomes and Deterioration
- Personality Disorders
- Placebo Effect
- Practice-Based Research and Practice Research Networks
- Psychodynamic Therapy (PDT)
- Resistance and Reactance
- Self-Reflection and Awareness
- Suicide and Crisis Intervention
- Termination
- Therapist Factors
- Training
- Transference and Countertransference
- Trauma and/or PTSD
- Treatment Length and Frequency
January 2024
How Can Psychotherapists Accurately Measure a Therapeutic Alliance Rupture?
Babl, A., Rubel, J., Gómez Penedo, J. M., Berger, T., Grosse Holtforth, M., & Eubanks, C. F. (2023). Can session-by-session changes in self-reported alliance scores serve as a measure of ruptures in the therapeutic alliance? Journal of Consulting and Clinical Psychology. Advance online publication. https://doi.org/10.1037/ccp0000861
The therapeutic alliance is a robust predictor of patient outcomes in psychotherapy. One can define the alliance as the collaborative agreement between patient and therapist on the goals of therapy (what is the desired outcome) and tasks of therapy (how therapy should proceed), and the relational bond between patient and therapist. Ruptures in the therapeutic alliance are breakdowns or tensions in the relationship between therapist and patient in any of the three core components of the alliance (agreement on goals and tasks, and the bond). If ruptures go unaddressed, they can lead to poor outcomes or patient dropout. Despite the practical importance of the alliance and of identifying alliance ruptures, therapists often feel at a loss when it comes to detecting a rupture. Ruptures can be characterized by patient behaviours that indicate withdrawal from the therapist (going silent, abstract, or vague responses) or confrontation with the therapist (complaining or rejecting interventions). However, therapists may not always be aware that a rupture in the alliance has occurred, and so they may need help to identify an alliance rupture. One source of help is to have patients complete a brief self-report measure of the state of the therapeutic alliance at the end of each session. In this study, Babl and colleagues asked 58 patients with depression who received CBT to rate the therapeutic alliance with a short validated scale (i.e., the Working Alliance Inventory [WAI]) after each of 20 sessions. The authors also video-recorded each session and rated alliance ruptures within each session with the 3RS a validated observer rating scale. Although the 3RS is the best way to assess alliance ruptures, 3RS ratings require an intensive process of coding by expert judges. And so, if the WAI, which is a relatively easy procedure, can reliably identify ruptures, then this may give clinicians a way of knowing when a rupture occurred. Rubl and colleagues defined an alliance rupture as a significant drop in the patient’s WAI score from one session to the next (i.e., technically, they used a score 2 standard deviations lower than the patient’s average WAI score in all preceding sessions). They found that sessions for which the WAI indicated a drop in alliance ratings had significantly more rupture markers (as rated by the 3RS) than sessions that did not show a drop in alliance ratings.
Practice Implications
Although somewhat preliminary, the results of this study suggest that clinicians can use changes in patient-self-reported WAI scores from session to session to identify if a therapeutic alliance rupture occurred. If there is a precipitous drop in WAI scores after a session, then therapists might enquire with the patient about the state of the therapeutic alliance in the next session. That is, therapists might ask if there is still an agreement on the goals of therapy or if the therapy is proceeding in a way that is congruent with what the patient expects. Alternatively, the therapist may check in on the state of the therapeutic relationship, that is, to assess if the patient may feel some tension, disappointment, or frustration with the therapist. The goal is for therapists to use relatively easy-to-use measures like the WAI to help them make the most of the therapeutic alliance.
Is CBT for Perfectionism Effective?
Smith, M. M., Hewitt, P. L., Sherry, S. B., Flett, G. L., Kealy, D., …, & Bakken, K. (2023, April 27). A meta-analytic test of the efficacy of cognitive behavioural therapy for perfectionism: A replication and extension. Canadian Psychology / Psychologie canadienne. Advance online publication. https://dx.doi.org/10.1037/cap0000360
Perfectionism is a multidimensional personality factor that can make one vulnerable to mental health problems and can interfere with therapeutic processes like the therapeutic alliance. The bulk of the research for treating perfectionism has focused on cognitive behavioural therapy (CBT), with meta-analyses concluding that CBT is efficacious. However, this research has certain limitations. First, the meta-analyses focus on standard mean differences (SMD) between treatments and control conditions, but SMDs are abstractions that are difficult to translate into real-world practice. What matters most to patients and clinicians is the proportion of patients who reliably improve beyond statistical error or the natural course of the problem. Second, the studies do not consistently report analyses related to drop-out rates. Drop-out is an indicator of the tolerability of a treatment for patients. If patients drop out at a high rate, then the treatment may not be acceptable, and many patients may not benefit. Third, the meta-analyses of CBT for perfectionism do not indicate whether treatment effects at the end of therapy are sustained in the longer term. In this re-analysis and extension of previous meta-analyses, Smith and colleagues look at the critical issues of the proportion of improved patients, drop-outs, and long-term outcomes for those with perfectionism treated with CBT. Like previous meta-analyses assessing SMDs, the authors found moderate to large effects of CBT versus control conditions on several of the dimensions of perfectionism. However, they also found no significant effects for several other dimensions of perfectionism that were more persistent and related to core personality (e.g., self-oriented perfectionism, other-oriented perfectionism, socially prescribed perfectionism, perfectionistic cognitions). Regarding the proportion of reliably improved patients, 13% to 55% of patients showed reliable improvement depending on the dimension of perfectionism that was assessed. But between 7.5% and 24.5% reliably improved without any treatment. So, the added value of CBT beyond no treatment was modest in most cases. Regarding dropouts, 27.0% of patients dropped out of CBT while only 15.7% dropped out of the control conditions. That is, those who received CBT had a 1.89 times increased risk of dropping out compared to those who were in control conditions (95% CI [1.37, 2.60], p = .005). Finally, very few studies assessed the longer-term effects of CBT, and those that did assess longer-term outcomes found that any effects at post-treatment disappeared at the follow-up assessments.
Practice Implications
The results of this meta-analysis suggest that it may not be enough to provide CBT to help patients who have the persistent relational aspects of perfectionism (i.e., self-oriented perfectionism, other-oriented perfectionism, socially prescribed perfectionism, and perfectionistic cognitions). It is also unclear if CBT is effective in the longer term as the small amount of evidence that exists does not inspire optimism. CBT may not be tolerable to some patients with perfectionism as evidenced by the moderately high drop-out rate. It may be necessary to adapt CBT or to include other models of treating perfectionism that use interpersonal and dynamic concepts and interventions that explicitly target core personality features of perfectionism.
Was Hans Eysenck Right About Psychotherapy’s Ineffectiveness?
Cuijpers, P., Karyotaki, E., Reijnders, M., & Ebert, D. (2019). Was Eysenck right after all? A reassessment of the effects of psychotherapy for adult depression. Epidemiology and Psychiatric Sciences, 28(1), 21-30. https://doi.org/10.1017/S2045796018000057
In the 1950s Hans Eysenck, a notorious figure in the history of psychology, suggested that psychotherapy was not effective. This set off a crisis of confidence in the psychotherapy world, but it also energized decades of research to assess whether Eysenck was right. Over the years that followed, many studies were published to assess the efficacy of psychotherapy, and methodologies like meta-analyses were developed specifically for this research area. Meta-analyses are a way of aggregating the effects of many studies in a research area so that one can estimate an average effect based on these studies. The first meta-analyses published in the 1970s showed that psychotherapy was indeed effective, but even back then there were doubts about the quality of the research. The “garbage in, garbage out” principle recognized that if the quality of studies were low, then the meta-analyses would produce biased results. Quality refers to studies that have sufficiently large samples, independent randomization of patients, blinded assessments of outcomes, equivalent training and supervision of all therapists, and all participant data included in the analyses. There is lots of evidence to show that low-quality studies overestimate the effects of treatments. Finally, there is the problem of publication bias, which occurs when researchers do not publish studies if the results are not supportive of the effectiveness of a treatment (i.e., if unpublished, these studies are not available to be included in a meta-analysis). In this meta-analysis, Cuijpers and colleagues focused on psychotherapies for depression because this is a large research area, depression is prevalent in the population, and there are sufficient studies that compare different forms of therapy. The unadjusted effect size of psychotherapy for depression was moderately large (g = .63 to .70), which suggests that about 4.18 patients need to be treated before one patient improves. But only 23% of 369 studies were considered to have a low risk of bias. When the authors accounted for risk of bias (e.g., they removed studies with high risk of bias and statistically adjusted for publication bias), the adjusted effect size of psychotherapy dropped to g = .31. This is a small effect, but still above the commonly accepted threshold of g = .24 indicating that a treatment is clinically relevant.
Practice Implications
Eysenck was probably wrong – psychotherapy as tested in clinical trials is effective and clinically relevant, but perhaps not as effective as once thought. Earlier estimates of psychotherapy’s effectiveness appear to have been inflated partly by biased studies conducted by enthusiastic adopters of newer therapies. The problem of bias also exists for other treatments as well, such as anti-depressant medications, so psychotherapy is not alone. Psychotherapists must strike a balance between being confident in what they are providing to patients and honestly assessing the efficacy of what they offer as treatment.
December 2023
Is empathic accuracy enough? The role of therapists’ interventions in the associations between empathic accuracy and session outcome.
Stolowicz-Melman, D., Lazarus, G., & Atzil-Slonim, D. (2023). Is empathic accuracy enough? The role of therapists’ interventions in the associations between empathic accuracy and session outcome. Journal of Counseling Psychology, 70(6), 682–690. https://doi.org/10.1037/cou0000708
Therapist empathy (as rated by patients) is a well-known predictor of positive patient outcomes in psychotherapy. Empathy can take many forms (empathic resonance, expressed empathy, received empathy), but one that is less well-researched is the therapist’s empathic accuracy. Empathic accuracy refers to the congruence (agreement) between a patient’s emotional state and the therapist’s perception of the patient’s emotional state. That is, therapists’ capacity to accurately perceive their patient’s inner experiences. In this study, Stolowicz-Melman and colleagues assessed therapist and patient congruence in therapists’ ratings of the patient’s emotional state during a session. The authors also looked at the impact of that congruence (or lack of congruence) on the patient’s ratings of the session’s effectiveness (session outcomes). In addition, the researchers examined whether different types of therapist interventions affected the impact of congruence (empathic accuracy) on a patient’s evaluation of a session. The authors hypothesized that more exploratory interventions (therapists identifying and labelling feelings, focusing on moment-to-moment client emotions) and fewer directive interventions (therapists setting an agenda, reviewing homework) would result in a stronger association between therapist empathic accuracy and session outcomes as rated by the patient. The study had 81 patients and 50 therapists. Patients and therapists rated the patient’s mood after each of at least 6 sessions of therapy. The congruence (agreement) between the patient's and therapist's rating of the client's mood was an index of therapist empathic accuracy. Clients also evaluated the effectiveness of each session of therapy. The results suggested that greater therapist empathic accuracy for negative emotions was associated with better session evaluations (outcomes) rated by patients. Exploratory, and not directive, interventions resulted in a stronger effect of therapist empathic accuracy for negative emotions on session evaluations.
Practice Implications
The results suggest that therapists need to recognize negative emotions in their patients, offer support and empathy, and at the same time help the patient explore and process these emotions. Empathic accuracy is not enough, however, when a patient is experiencing a negative emotion. Patients need therapists to help them recognize their experiences and to explore their meaning in words. As Stolowicz-Melman and colleagues conclude, therapist exploration of patients’ emotions may help patients tolerate and regulate painful feelings and achieve better session results.
Psychotherapies for the treatment of borderline personality disorder: A systematic review.
Crotty, K., Viswanathan, M., Kennedy, S., Edlund, M. J., Ali, R., Siddiqui, M., Wines, R., Ratajczak, P., & Gartlehner, G. (2023, October 30). Psychotherapies for the treatment of borderline personality disorder: A systematic review. Journal of Consulting and Clinical Psychology. Advance online publication. https://dx.doi.org/10.1037/ccp0000833
Borderline personality disorder (BPD) is the most common of the personality disorders affecting about 1.8% of the population. BPD is particularly present in psychiatric outpatients (10%) and inpatients (15% to 25%). Those with BPD have severe functional impairment caused in part by high rates of comorbid mental disorders, substance use, self-harm, and suicidal ideation. Clinical practice guidelines recommend psychological therapies as the first-line treatment approach, with a particular emphasis on dialectical behaviour therapy (DBT). However, recent research shows that such treatment guidelines often do not rely on a systematic review of the literature that includes an analysis of the quality of the evidence. Low-quality evidence may not be reliable and lead to erroneous conclusions and recommendations. In this systematic review, Crotty and colleagues provide an update to clinical practice guidelines for BPD considering the quality of the evidence. The authors found 25 studies (20 of which were randomized controlled trials) of individuals 13 years and older diagnosed with BPD. These studies represented 2,545 participants in which a psychological treatment was compared to another treatment or treatment as usual (TAU). The most studied treatment in 6 studies was DBT. Other treatments included mentalization-based treatment, transference-focused psychotherapy, schema therapy, systems training for emotional predictability and problem-solving, dynamic deconstructive therapy, acceptance and commitment therapy, and CBT. Only 1 of the 25 studies had a low risk of bias (i.e., had no or few problems with randomization, data analysis, reporting of dropouts, sample size, blinding of assessment, etc.). Overall, TAU and the nine psychotherapies tested were effective in reducing the severity and symptoms of BPD. The authors did not find consistent evidence that DBT was superior to other psychotherapies.
Practice Implications
Psychological treatments in general are effective in reducing symptoms of BPD. However, the quality of the evidence is not high. Practice guidelines focus on DBT as a treatment for BPD, but this systematic review as well as previous meta-analyses, did not find that DBT was superior or that any psychotherapy is more beneficial than another. The authors concluded that generalized psychotherapies may be as effective as intensive specialized treatments for BPD that require expert training.
Five decades of research on psychological treatments of depression: A historical and meta-analytic overview.
Cuijpers, P., Harrer, M., Miguel, C., Ciharova, M., & Karyotaki, E. (2023, November 16). Five decades of research on psychological treatments of depression: A historical and meta-analytic overview. American Psychologist. Advance online publication. https://dx.doi.org/10.1037/amp0001250
There is not much new to say about the effects of psychotherapy on depression. Numerous meta-analyses of a large body of evidence have been published over the past few years. Meta-analysis is a way of combining the effects of many studies to produce an overall average effect, and the procedure also allows one to examine the reliability of the findings and predictors of the effect sizes. Meta-analyses are not perfect, but compared to interpreting a single study meat-analyses provide us with a more reliable sense of what is happening in a research area. Research has demonstrated that several types of psychotherapy are equally effective in treating depression including CBT, IPT, third-wave therapies, psychodynamic therapy, and nondirective counselling. Psychotherapy has comparable effects to antidepressant medications in the short term, but psychotherapy has better effects in the longer term. The effects of psychotherapy for depression are comparable across different groups (women with perinatal depression, older adults, and those with medical conditions), but psychotherapies have smaller effects in children and adolescents and those with comorbid substance use disorders. In this overview, Cuijpers and colleagues looked at trends in this research over the past five decades. They meta-analyzed 562 studies of psychological treatments for depression. They found that almost half of the trials were of CBT, and 8 other therapies made up the other half of the trials. The number of studies per year is increasing with an average of 8.6 studies published per year. Only 34% of studies met the criteria for low risk of bias – that is, 66% of studies had some problems with random allocation, blinding of assessors, data analysis, sample size, etc. We know that a high risk of bias inflates effect sizes, and so it is likely that the effects of many studied treatments are inflated. Unfortunately, the relative number of studies testing therapies other than CBT seems to be declining over the years. Trials in Europe are increasing relative to studies in North America.
Practice Implications
This meta-analysis is consistent with the many that preceded it. Psychological treatment for depression is modestly effective, there are no differences between treatment types, although, in the long run, psychotherapy is more effective than antidepressant medications. What is concerning is that research on new treatments may be declining which may further limit the information and options that therapists and patients have about evidence-based treatments for depression.