Blog

McLaughlin, S.P.B., Barkowski, S., Burlingame, G.M., Strauss, B., & Rosendahl, J. (2019). Group psychotherapy for borderline personality disorder: A meta-Analysis of randomized-controlled trials. Psychotherapy. Advance online publication. http://dx.doi.org/10.1037/pst0000211

Borderline personality disorder (BPD) is characterized by fear of abandonment, unstable intense relationships, rapid changes in identity and self-image, impulsivity, wide mood swings, periods of intense anger, and ongoing feelings of emptiness. These symptoms sometimes lead to suicidal behavior or non-suicidal self-injury. Often, those with BPD report a very stressful childhood that included sexual and/or physical abuse, and neglect. Borderline personality disorder is the most common of the personality disorders and is associated with severe social psychological impairment such that those with BPD often have unstable employment, are involved in abusive relationships, and engage in risky behaviors. A diagnosis of BPD is also associated with a high rate of mortality due to suicide. Practice guidelines indicate that psychotherapy is a key component to the treatment of BPD. Two psychological treatment approaches that incorporate group interventions are dialectical behavior therapy (DBT) and mentalization-based treatment (MBT). In DBT patients learn specific skills to alter maladaptive ways of regulating emotions in a group context. In MBT, an attachment-based treatment, the focus is on building trust in others through group interactions that generalize to other social relationships. In this meta-analysis, McLaughlin and colleagues reviewed 24 studies that compared group treatment for BPD to treatment as usual, which included a variety of interventions like supportive groups, pharmacotherapy, individual therapy, and others. Some of the treatments for BPD were stand-alone groups and some groups were part of a larger comprehensive program. Participants attended between 12 and 130 sessions, and group size ranged from 4 to 12 members. The meta-analysis revealed that group treatment for BPD versus treatment as usual resulted in moderate to large effect on BPD symptoms: g = .72, CI: [.41, 1.04], p < .001. The effects of group treatment versus treatment as usual on suicidality produced a moderate effect, g = .46, CI: [.22, .71], p < .001. The authors reported similar results for secondary outcomes like depression, anxiety, and general mental health. Drop-out rates were similar between group treatments (26.26%) and treatment as usual (28.26%). There were no differences in the effects of group therapy orientations on any of the outcomes or on drop-out rates.

Practice Implications

The results of this meta-analysis indicated the value of group treatment for BPD not only for core symptoms and suicidality, but also for symptoms related to quality of life (depression, anxiety). Theoretical orientation did not explain any of the findings, suggesting that treatments like DBT and MBT in a group format are equally effective. Therapists and patients can feel confident that group treatment for BPD are among the most effective treatments available.

Author email: mclaughlin.stevie@gmail.com

Carl, E., Witcraft, S.M., Kauffman, B.Y., Gillespie, E.M., Becker, E.S…. Powers, M.B. (2019). Psychological and pharmacological treatments for generalized anxiety disorder (GAD): a meta-analysis of randomized controlled trials. Cognitive Behaviour Therapy, DOI:10.1080/16506073.2018.1560358

Generalized anxiety disorder (GAD) is characterized by excessive and difficult to control worry about events or activities. GAD is associated with a high level of impairment in social functioning, work productivity, and health-related quality of life. GAD is also associated with a high level of medical costs and health care utilization. About 4.3% of the general population have experienced GAD at one time in their life. In this updated meta-analysis, Carl and colleagues reviewed the empirical literature to compare the effects of psychotherapies and pharmacotherapy to control conditions. Seventy-nine studies with over 11,000 participants were included in the review. In 39 comparisons, evidence-based psychotherapies outperformed control conditions on measures of anxiety at posttreatment (g = 0.76, 95% CI: 0.61–0.91, p < 0.001), suggesting a medium to large effect. Only 12 studies evaluated follow-up data, and they found that psychotherapy resulted in a small but statistically significant average effect on anxiety symptoms (g = 0.27, 95% CI: 0.00–0.53, p = 0.05). Compared to older patients, younger patients tended to do better in psychotherapy. Forty-three studies found that pharmacotherapy consistently outperformed control conditions at post-treatment (g = 0.38, 95% CI: 0.30–0.47, p < 0.001) suggesting a small effect. There were no studies that assessed pharmacotherapy at a follow-up date. Patient age or treatment dose did not affect outcomes of pharmacotherapy. The authors were careful to point out that that the effect sizes of psychotherapy and pharmacotherapy were not comparable in this meta-analysis because psychotherapy trials tended to use no-treatment controls whereas pharmacotherapy trials tended to use placebo controls, and the latter tends to produce more conservative (smaller) estimates of effects.

Practice Implications

Both psychotherapy and pharmacotherapy appear to be effective by post-treatment for patients with GAD. The effects of psychotherapy at follow-up is diminished, and no studies evaluated whether patients receiving pharmacotherapy maintained any gains at follow-up. Research has suggested that compared to psychotherapy, pharmacotherapy outcomes for depression at follow up is poorer. Although this study does not allow one to compare psychotherapy to pharmacotherapy, evidence from another meta-analysis suggests that patients would strongly prefer psychotherapy if given the choice. And patients receiving their preferred treatment tend to experience significantly better outcomes.

Author email: emilycarl@utexas.edu

Norcross, J.C. & Wampold, B.E. (2018). A new therapy for each patient: Evidence‐based relationships and responsiveness. Journal of Clinical Psychology, Online First, DOI: 10.1002/jclp.22678

Over the next several months, I will review in this blog results of a number of meta-analyses conducted recently on patient factors and relationship factors in psychotherapy. These factors provide evidence-based guidance to psychotherapists on how best to relate to and adapt to clients so that psychotherapy is more effective. This introductory article by Norcross and Wampold is an overview of the nine meta analyses related to transdiagnostic client factors to which therapists can adapt their interpersonal stances and treatment. The goal is to enhance treatment effectiveness by therapists tailoring therapy to individual client characteristics that are related to outcomes. Decades of research indicate that client transdiagnostic characteristics have more influence on outcomes than the particular treatment method, and likely more influence than the particular client diagnosis. The research indicates that giving the identical treatment to every client without adaptation to client characteristics is not an effective approach to providing psychotherapy. These meta analyses of client factors indicate that therapists should select different interventions and relational stances according to the client and the context. What are these client characteristics and therapist adaptations that are reliably related to outcomes? The client factors most strongly related to outcomes include therapist adaptations to: client culture/race/ethnicity (99 studies, g = .50); client preferences for type of therapy (51 studies, g = .28), client religion/spirituality (97 studies, g = .13 to .43), client stage of change (76 studies, g = .41), client reactance/resistance level (13 studies, g = .78), client coping style (32 studies, g = .53), and client attachment style (32 studies, g = .35). Over the next months, I will be reviewing in more detail these meta analyses of client factors and the practice implications of each so that therapists can use this evidence-base to help them to adapt to particular client characteristics.

Practice Implications

Practitioners will find that fitting the therapy to clients’ culture, stage of change, religion/spirituality, reactance/resistance, coping style, and attachment style will improve treatment outcomes. Doing so will have a greater impact on outcomes than the particular type of therapy provided or adapting treatment to the particular client diagnosis. The results of this large body of evidence suggests that therapists should no longer ask: “what is my theoretical orientation” but rather they should ask: “what relationship, adaptation, and approach will be most effective with this particular client”.

Norcross, J. C., & Lambert, M. J. (2018). Psychotherapy relationships that work III. Psychotherapy, 55(4), 303-315.

Relationship factors in psychotherapy are some of the most important predictors of patient outcomes. They outweigh factors like the type of therapy provided in determining whether patients get better after psychotherapy. In this second overview article, Norcross and Lambert provide a review of 17 meta-analyses of relationship factors in psychotherapy that contribute to positive outcomes. Like the review of patient factors also found in this blog and E-Newsletter, this article briefly outlines those evidence-based relationship factors that reliably predict patient outcomes in psychotherapy. The therapeutic relationship refers to how the therapist and patient relate to each other, or their interpersonal behaviors. By contrast, techniques or interventions refer to what is done by the therapist. Practice guidelines typically focus on interventions or therapeutic orientation. As the authors argue, what is missing from treatment guidelines are the person of the therapist and the therapeutic relationship – evidence for which is backed up by 5 decades of research. Even in studies of highly structured manualized psychotherapy for a specific disorder in which efforts were made to reduce the effect of individual therapist, up to 18% of outcomes (a moderate to large effect) could be attributed to the person of the therapist. By contrast somewhere between 0% and 10% of outcomes (a small to moderate effect) is attributable to specific treatment methods. So, which therapeutic relationship factors are reliably related to patient outcomes? These include: the therapeutic alliance in individual therapy (306 studies, g = .57) couple therapy (40 studies, g = .62), and adolescent psychotherapy (43 studies, g = .40), collaboration (53 studies, g = .61) and goal consensus (54 studies, g = .49), cohesion in group therapy (55 studies, g = .56), therapist empathy (82 studies, g = .58), collecting and delivering client feedback or progress monitoring (24 studies, g = .14 to .49), managing countertransference (9 studies, g = .84), and repairing therapeutic alliance ruptures (11 studies, g = .62) among others. Over the next few months, I will be reviewing these meta analyses in more detail to discuss how therapists can use this evidence base to improve their patients’ outcomes.

Practice Implications

The research as a whole indicates that therapists should make the creation and cultivation of the therapeutic relationship a primary goal of therapy. Factors such as managing the therapeutic alliance, repairing alliance ruptures, engaging in ongoing progress monitoring, managing countertransference and others should be used to modify treatments and interpersonal stances in order to maximize outcomes. When seeking out professional development and training, practitioners should focus on evidence-based relationship factors (managing the alliance, judicious self disclosure, managing emotional expression, promoting credibility of the treatment, collecting formal feedback, managing countertransference) in addition to focusing on evidence-based treatments.

Grenon, R., Carlucci, S., Brugnera, A., Schwartze, D., … Tasca, G. A. (2018). Psychotherapy for eating disorders: A meta-analysis of direct comparisons, Psychotherapy Research, DOI: 10.1080/10503307.2018.1489162

Eating disorders can cause a great deal of physical and mental impairment because of the severity of the symptoms and because of comorbid conditions like depression, anxiety, substance use, and others. Anorexia nervosa (AN) occurs in about 0.5% of the population, bulimia nervosa (BN) occurs in about 1.5% of the population, and binge-eating disorder (BED) occurs in about 3.5% of the population. Treatment guidelines include both cognitive behavioral therapy (CBT) and interpersonal psychotherapy (IPT) as front line interventions for BN and BED. However, results from previous meta analyses of psychological treatments for eating disorders were confounded by not focusing exclusively on randomized controlled trials, mixing studies of adult and adolescent samples, combining an array of outcomes rather than separately reporting primary (eating disorder symptoms) and secondary (interpersonal problems, depression) outcomes, and not distinguishing between bona fide psychotherapies (like CBT, IPT, psychodynamic therapy, and others) from non-bona fide treatments (like self help, behavioral weight loss supportive counseling). Grenon and colleagues conducted a meta analysis of psychotherapies for eating disorders to examine if: psychotherapy is effective compared to a wait list, if bona fide psychotherapy and non-bona fide treatment differ in outcomes, and if one type of psychotherapy (i.e., CBT) was more effective than other bona fide psychotherapies (like IPT, behavior therapy, psychodynamic therapy, dialectical behavior therapy). Their meta analysis included 35 randomized controlled trials of direct comparisons. Psychotherapy was significantly more effective than a wait-list control at post treatment, so that 53.89% of patients were abstinent of symptoms after psychotherapy compared to only 8.92% who were abstinent in the wait-list group. Bona fide psychotherapies (51% abstinent) were significantly more effective than non-bona fide treatments (40% abstinent) at post treatment, and dropout in bona fide psychotherapies (17.5%) was significantly lower than in non-bona fide treatment (29.1%). Further, the difference between CBT and other bona fide psychotherapies was not significant.

Practice Implications

Psychotherapy for eating disorders are effective for patients with BN or BED. There were too few studies of those with AN to come to any conclusions about their treatment. Patients with BN or BED are best treated with a bona fide psychotherapy that involves face to face psychological therapy like CBT, IPT, psychodynamic therapy, dialectical behavior therapy, or behavior therapy. Non-bona fide treatments like self help, behavioral weight loss, and supportive counseling should only be used as an adjunct to bona fide psychotherapy for eating disorders.

Flückiger, C., Del Re, A. C., Wampold, B. E., & Horvath, A. O. (2018). The alliance in adult psychotherapy: A meta-analytic synthesis. Psychotherapy. Advance online publication. http://dx.doi.org/10.1037/pst0000172

The therapeutic alliance is one of the most researched concepts in psychotherapy. The alliance, also called the working alliance or therapeutic alliance, consists of the collaborative agreement between patient and therapist on the tasks (what to do) and goals (what to achieve) of their therapeutic work together. Alliance also includes the relational or emotional bond between therapist and patient. It is different from therapist empathy, transference, countertransference, the real relationship and other concepts related to the therapeutic relationship. Researchers and clinicians have known for years about the importance of developing and maintaining an alliance to achieving patient outcomes. The growing research in this area now allows one to see how stable this finding is. Fluckiger and colleagues conducted a meta analysis of 306 studies with over 30,000 patients that assessed the alliance-outcome relationship. The research occurred in naturalistic settings (during regular clinical practice) and in randomized controlled trials. The overall effect size based on 295 independent comparisons was r = .278 (95% CI: .256, .299), indicating a statistically significant medium-sized association accounting for about 8% of treatment outcomes. To put this in perspective, this effect is as large as or larger than the effects of many common medical interventions. The type of therapy made no difference to this finding - the alliance was just as important to CBT as it was to psychodynamic, interpersonal, and emotionally focused therapies. The alliance-outcome correlation was somewhat smaller, though still significant among those with substance-use disorders, but otherwise was consistent for all other disorders tested (depression, anxiety, PTSD, borderline personality disorder). The alliance measure used, who rated the alliance, when it was assessed, and the outcome that was measured tended to have a small or no impact on the results. The alliance-outcome relationship was just as important to everyday clinical practice as it was in randomized controlled trials.

Practice Implications

The alliance-outcome association is highly reliable or stable across a number of therapies, diagnoses, measurements, and study designs. This very large body of research suggests that therapists should: (1) build and maintain an emotional bond, and agreement on tasks and goals with patients throughout therapy; (2) develop the alliance early by focusing on agreement on treatment and goals; (3) address ruptures in the alliance early and immediately; and (4) assess the strength and quality of the alliance regularly throughout treatment from the patient’s perspective using a well-known brief alliance measure.