Blog

Cuijpers, P., Driessen, E., Hollon, S. D., van Oppen, P., Barth, J., & Andersson, G. (2012). The efficacy of non-directive supportive therapy for adult depression: a meta-analysis. Clinical psychology review, 32(4), 280-291.

The research evidence indicates that there is very little difference between different types of psychotherapy (CBT, IPT, PDT, EFT, and others) in patient outcomes, especially for depression. Nondirective supportive treatment (NDST) also shows positive outcomes for various disorders. NDST is often used as a “placebo” condition in psychotherapy trials to control for common or non-specific factors. Common factors refer to those aspects that are common to all therapies, but that are not specific to any one therapy (e.g., therapist interpersonal skills, therapeutic alliance, client expectations). NDST does not involve specific therapeutic interventions like cognitive restructuring, transference interpretations, two-chair techniques, etc. In this meta analysis, Cuijpers and colleagues assessed those randomized controlled trials for depression in which specific treatments (e.g., CBT, PDT, IPT, EFT) or no treatment control conditions were directly compared to NDST. By doing so, the authors were able to estimate how much of patient outcomes were attributable to: specific effects of treatments (the difference between a specific intervention and NDST), common effects of treatment (the difference between NDST and no treatment), and extra-therapeutic factors (the effects of no treatment). The meta analysis included 31 studies with over 2500 patients with depression. Twenty-one comparisons included CBT, and the rest included IPT, PDT, or EFT. NDST was significantly less effective than other specific therapies (e.g., CBT, IPT, PDT, or EFT) at post-treatments g = −0.20 (95% CI: −0.32 to −0.08), but the effect was quite small. The difference between NDST and CBT alone (the most researched treatment type) was not statistically significant. Interestingly, when the authors controlled for researcher allegiance (an indication of which treatment was preferred by the researcher), the superior effects of specific treatments over NDST disappeared. NDST was significantly more effective than no-treatment, and the effect was moderate, g=0.58 (95% CI: 0.45–0.72). Pre- to post-treatment change in symptoms in the control condition was statistically significant, g = 0.39 (95% CI: 0.03–0.74), indicating the positive effects of extra-therapeutic factors on depressive symptoms (e.g., events in the patient’s life not related to therapy). Overall, the authors were able to estimate that almost 50% of patient outcomes could be attributed to common factors (therapist interpersonal skills, therapeutic alliance, client expectations, etc.), about 17% was due to specific therapy techniques (cognitive restructuring, two chair techniques, IPT interventions), and about 33% was due to extra-therapeutic factors (e.g., the natural course of depressive symptoms or other events in the patient’s life).

Practical Implications

Factors like therapist interpersonal skills and managing the therapeutic relationship appear to account for most (50%) of why patients with depression get better. The specific interventions based on therapy models like CBT account for relatively less of patient outcomes (17%). The natural course of the disorder and other events in patients’ lives account for about a third of patient improvement. Therapists can learn how to maximize the effects of common factor skills through deliberate practice and training to identify and repair alliance ruptures to help their patients get better.

So, M., Yamaguchi, S., Hashimoto, S., Sado, M., Furukawa, T.A., & McCrone, P. (2013). Is computerised CBT really helpful for adult depression?-A meta-analytic re-evaluation of CCBT for adult depression in terms of clinical implementation and methodological validity. BMC Psychiatry, 13, 113. 

Depression is a major cause of disability in the world, and so efforts to improve access to its treatment have been ongoing for several decades. In particular, many researchers and clinicians propose cognitive behavioural therapy (CBT) as an effective treatment with a good evidence-base. There have been many clinical trials showing the efficacy of CBT. In recent years, there have also been attempts to computerize CBT (CCBT) as a self help intervention in order to increase its accessibility for those with depression, and perhaps also to improve its cost effectiveness. In fact, the Increasing Accessibility to Psychotherapy (IAPT) program in the UK provides CCBT as the most common first treatment for depression. However there remain questions about the longer term effectiveness of CCBT to reduce symptoms of depression, its potentially high patient dropout rate (a negative outcome), and its effects on quality of life of those burdened by depression. In this meta analysis, the largest of its kind, So and colleagues assess these issues with regard to CCBT. They reviewed 14 direct comparison randomized controlled trials that provided 16 comparisons of CCBT versus a control condition (wait list or treatment as usual) for adults with depression. At post-treatment, CCBT was more effective than controls in reducing depression −0.48 [95% CI −0.63 to −0.33]. However, at follow up (up to 6 months), the effects of CCBT disappeared −0.05 [95% CI −0.19 to 0.09]. Also improvement in functioning and quality of life were not significantly different between CCBT and control conditions, −0.05 [95% CI −0.31 to 0.22]. The rate of drop out from CCBT (32%) was almost double that of control conditions (17%), RR = 1.68 [95% CI 1.31 to 2.16]. There was also evidence of publication bias (i.e., a tendency for some researchers not to publish non-significant findings), so that the positive post-treatment results in favour of CCBT might be inflated.

Practice Implications

Although CCBT may be touted as a way to increase access to treatment for depression, this meta analysis indicates some concerns about the widespread implementation of CCBT. The effects of CCBT appear to be limited to a short-term reduction of depressive symptoms that may not be sustained in the longer run. There was no appreciable impact of CCBT on quality of life relative to controls, and so CCBT may have a limited impact on the burden of depression. Most troubling was a high drop out rate of 32%. Drop out from CCBT in the IAPT program in the UK is about 50%, and this may be indicative of the actual drop out rate in real world practice.

Gutermann, J., Schwartzkopff, L., & Steil, R. (2017). Meta-analysis of the long-term treatment effects of psychological interventions in youth with PTSD symptoms. Clinical Child and Family Psychology Review, 20, 422-434.

Natural disasters, physical abuse, sexual abuse, war, accidents, loss and severe illness are traumatic events that can occur during childhood and adolescence. These potentially traumatic events are highly prevalent in youth, and approximately 15% of children and adolescents who have been exposed to traumatic events meet the diagnostic criteria for post-traumatic stress disorder (PTSD). Criteria for a diagnosis of PTSD include: intrusive memories of the traumatic event, avoidance, hyperarousal, and negative change in mood or cognitions. PTSD symptoms are also highly stable over time, and so without intervention they do not tend to improve. In this meta-analysis, Gutermann and colleagues assess the effects of psychological treatments for PTSD in youth, with a special emphasis on their long term therapeutic effects. Forty-seven studies of 3767 participants were included in the analyses. Traumas were varied and included childhood abuse, physical abuse, accidents, wars, and natural disasters. About 68% of interventions were CBT-oriented, and 67% were provided in a group therapy format. The uncontrolled pre-treatment to follow-up effect sizes for PTSD symptoms was large for studies with a follow-up period greater than 6 months (N = 30; g = .99, CI .83, 1.16). However, when psychological interventions were compared to treatment as usual or an active control group in a randomized controlled trial, the effects at post-treatment were small (N = 6; g = .38, CI .03–.74), and effects at follow up periods combined were also small (N = 19; g = .38, CI .20, .55).

Practice Implications

Psychological interventions resulted large effects to reduce PTSD symptoms from pre-treatment to follow-up from treatment. However, compared to treatment as usual or other active control groups, psychological treatments resulted in small effects in the longer term. There were too few studies to assess different treatment approaches, age groups, and modalities (group vs individual). Nevertheless, the results provide support for the efficacy of psychological treatments for PTSD in youth with modest effects at follow-up.

Author email: Gutermann@psych.uni-frankfurt.de

Grenon, R., Schwartze, D., Hammond, N., Ivanova, I., Mcquaid, N., Proulx, G., & Tasca, G. A. (2017). Group psychotherapy for eating disorders: A meta-analysis. International Journal of Eating Disorders. DOI: 10.1002/eat.22744

Group therapy has an evidence base indicating its efficacy for many disorders. Groups represent a social microcosm in which interpersonal factors that underlie psychological distress and symptoms can be effectively addressed. Group therapeutic factors include peer interpersonal feedback, social learning, emotional expression, and group cohesion. Theories of eating disorder symptoms include interpersonal problems and affect dysregulation as maintenance factors. Many treatment guidelines indicate that individual and group CBT are the treatments of choice for eating disorders. However, there are no meta analyses that specifically look at the efficacy of group therapy for eating disorders. In this study, Grenon and colleagues assess if: (a) group psychotherapy for eating disorders is efficacious compared to wait-list controls, (b) group therapy is effective compared to other active treatments (self help, individual therapy, medications), and (c) group CBT is more effective than other types of group therapy (group interpersonal therapy [GIPT], group psychodynamic-interpersonal psychotherapy [GPIP], or group dialectical behavior therapy [GDBT]). The authors reviewed 27 randomized controlled trials with over 1800 patients that provided direct comparisons of group therapy for eating disorders. The mean drop out rate from group therapy was 16.47% (SD = 13.46), which is similar to what is reported for psychotherapy trials in general. Group therapy was significantly more effective than wait list controls in achieving abstinence from binge eating and purging (RR = 5.51, 95% CI: 3.73, 8.12), decreasing the frequency of binge eating and/or purging (g = 0.70, 95% CI: 0.51, 0.90), and reducing related psychopathology (g = 0.49, 95% CI: 0.32, 0.66). Group psychotherapy had an overall rate of abstinence from binge eating of 51.38%, while wait-list control conditions had an overall abstinence rate of 6.51%. Similar findings were achieved a follow-ups. The effects of group psychotherapy and other active treatments (e.g., behavioral weight loss, self-help, individual psychotherapy) did not differ on any outcome at post-treatment or at follow-ups. Group CBT and other forms of group psychotherapy did not differ significantly on outcomes at any time point.

Practice Implications

The results add to a growing body of research that indicates that group psychotherapy is as effective as other treatments, including individual therapy, to treat mental disorders. Despite the fact that practice guidelines indicate that CBT is the treatment of choice for eating disorders, this meta analysis did not provide evidence that group CBT was more effective than other types of group treatments. Clinicians considering group interventions for eating disorders or other mental health problems will do well to make use of group therapeutic factors like interpersonal learning, peer feedback, emotional expression, and group cohesion to improve patient outcomes.

Schwartze, D., Barkowski, S., Strauss, B., Burlingame, G., Barth, J., & Rosendahl, J. (2017). Efficacy of group therapy for panic disorder: Meta-analysis of randomized controlled trials. Group Dynamics, 21, 77-93.

Panic disorder (PD) is characterized by recurrent episodes of intense fear or discomfort accompanied by physical and cognitive symptoms that may include sweating, trembling, or fear of dying. The panic attacks can lead to avoidant behavior that results in isolation, impaired functioning and lower quality of life. Often, those with PD also experience agoraphobia or an intense fear of having a panic attack in public, open spaces, or in a crowd. PD has a lifetime prevalence of 5% among adults in the US. Patients with PD use health care services at a higher rate than the general population, and those with PD may not receive adequate treatment. An evidence-based treatment for PD is cognitive behavioral therapy (CBT). Practice guidelines for PD recommend pharmacotherapy and psychotherapy with CBT. However, these practice guidelines do not take into account group therapy for PD. In this meta analysis, Schwartze and colleagues included group treatment studies of PD that were randomized controlled trials (RCT) and in which direct comparisons of group therapy to other treatments were conducted. RCTs of direct comparisons provide the best quality evidence of the efficacy of a treatment approach. The authors included 15 studies (14 of which were of group CBT for panic) that had 864 patients. There was a large significant effect on panic and agoraphobic symptoms favoring group over no-treatment controls (k = 9; g = 1.08; 95% CI [0.82, 1.34]; p = .001). Similar results were found for depressive symptoms and general anxiety symptoms. There was no significant difference between group and alternative PD treatments (pharmacotherapy, individual therapy) on the primary outcomes (k = 6; g = 0.18; 95% CI [-0.14, 0.49]; p = .264). Again similar results were found for depression and anxiety symptoms. In total 78% of patients with PD were symptom-free after group psychotherapy, compared with 33% in no-treatment control groups, and 71% in alternative treatment.

Practice Implications

The number of studies were small, but the results of this meta analysis indicate that group therapy is an effective treatment for PD and perhaps as effective as typical alternatives like pharmacotherapy and individual therapy. Group CBT protocols usually involve multiple components such as (a) education regarding the etiology and maintenance of PD, (b) cognitive restructuring (identifying and modifying panic-related cognitions), (c) exposure to external situations (in vivo exposure) or internal bodily sensations (interoceptive exposure), (d) relaxation training and/or breathing retraining. Group therapy may also provide a lower cost, more accessible, and possibly as effective treatment alternative than individual therapy for PD.

Driessen, E., Hollon, S.D., Bockting, C.L.H., Cuijpers, P., Turner, E.H. (2015). Does publication bias inflate the apparent efficacy of psychological treatment for major depressive disorder? A systematic review and meta-analysis of US National Institutes of Health-funded trials. PLoS ONE 10(9): e0137864. doi:10.1371/journal.pone.0137864.

In 2008 Turner published a well-known study in which he found that almost 50% of antidepressant trials registered with the Food and Drug Administration in the US were never published or were positively “spun” (i.e., essentially negative findings were interpreted to be positive). Almost all of the unpublished trials showed unfavorable results for the antidepressants’ effects. By contrast, the published studies were almost always were positive. This is evidence of publication bias caused by selective publication of some data and suppression of other data. As you can imagine, this has important implications for treatment of depression as the published record appeared to over-inflate effects of antidepressants by 25% (the mean effect size decreased from g = .41 [CI_95% 0.36~0.45] to 0.31 [0.27~0.35] when unpublished studies were included). Has the same type of publication bias occurred in the published record of psychotherapy’s efficacy? In this study by Driessen and colleagues, the authors reviewed all psychotherapy studies for depression funded by the National Institutes of Mental Health in the US between 1972 and 2008. They wanted to determine which ones were published, which were never published, and what the impact of nonpublication was on the mean effect size. Of the 55 grants that were funded, 13 (26.3%) were never published, and the authors were able to obtain data from 11 of those unpublished studies. The overall mean effect size (psychological treatment versus a control condition) of unpublished studies was g = 0.20 (CI_95% -0.11~0.51) indicating a small non-significant effect. The overall mean effect size for published studies was g = 0.52 (CI_95% 0.37~0.68) indicating a medium significant effect. Adding the unpublished studies to published studies resulted in a 25% decrease in effect size estimate to g = 0.39 (0.08~0.70), indicating a small but significant effect of psychotherapy.

Practice Implications

This study indicated that psychotherapy is effective but that the effects are likely smaller than indicated in the published record. As in the case of antidepressant medication research, a minority of researchers may not publish findings that are not in line with their preconceived expectations or wished-for results. Regardless, there is certainly room for psychotherapy to improve. After decades of focusing largely on the efficacy of specific psychotherapies like CBT, psychodynamic therapies, and interpersonal therapy, perhaps it is time to shift to studying how and why therapies work, and which patients benefit from specific interventions. There are promising avenues such as research on: repairing therapeutic alliance tensions, enhancing therapist expertise, progress monitoring and feedback, client factors, and managing countertransference.