Blog

The Great Psychotherapy Debate: Since in April, 2015 I review parts of The Great Psychotherapy Debate (Wampold & Imel, 2015) in the PPRNet Blog. This is the second edition of a landmark, and sometimes controversial, book that surveys the evidence for what makes psychotherapy work. You can view parts of the book in Google Books.

Wampold, B.E. & Imel, Z.E. (2015). The great psychotherapy debate: The evidence for what makes psychotherapy work (2^nd edition). New York: Routledge.

Why We Should Care About Allegiance Effects in Psychotherapy Research

Allegiance in psychotherapy refers to the degree to which a researcher or therapist believes that the therapy they are studying or delivering is effective. Clients have an expectation that therapists have an explanation for their disorder and that the therapy used to address that explanation will lead to improvements. On their part, psychotherapists choose a therapeutic approach that is consistent with their understanding of psychological distress. Wampold and Imel argue that therapist allegiance is a common factor across therapies that contributes to good patient outcomes. Although allegiance is an important therapeutic factor, it complicates the conduct of psychotherapy research. In a trial comparing two treatments, for example, researchers and therapists tend to be affiliated with one of the treatments and so they believe in the effectiveness of their treatment. They often do not feel the same way about the comparison treatment, or they may desire that their preferred treatment be more effective than the comparison. In medication trials, this allegiance effect can be controlled by a double blind placebo controlled design in which both therapist and patient are not aware of who is receiving which active medication, or who is receiving a placebo. It is impossible to blind therapists in psychotherapy trials – therapists have to know what treatment they are providing. When doing a meta analytic review of psychotherapy trials, it is possible and relatively easy to identify the allegiance of the researchers in a particular study by looking at their past publications, and by reading what they say about the therapies they are comparing. Often, the developer of a treatment manual is a co-author of the trial. Wampold and Imel review several meta analyses that assess the allegiance effects. In three meta analyses from published in 1980, 1999, and 2013 the correlation between ratings of researcher allegiance and effects of psychotherapy on patient outcomes ranged from moderate to large (r = .26 to r = .85). One interesting meta analysis illustrates the magnitude of this effect. The reviewers looked at 69 studies on self statement monitoring (SSM), a type of cognitive therapy developed by Meichenbaum. The average effect of SSM compared to controls in all studies was d = .53 to d = .74, which is moderate. However, effect sizes found in the studies co-authored by Meichenbaum were nearly twice as large, d = d = 1.23. Being a co-investigator in a study of a therapy that one develops, apparently doubles the effect of the treatment on patient outcomes.

Practice Implications

Therapist allegiance to a treatment is important to the effectiveness of the treatment in that therapist allegiance increases the therapist’s confidence in the treatment’s effectiveness and increases a patient’s expectation of getting better. However, when interpreting psychotherapy trials, especially those that pit one type of therapy against another, it is important to keep in mind the researchers’ allegiance. It is rare to see trials that compare two interventions in which the research team is made of up proponents of the two interventions. However such trials are important and necessary.

Markowitz, J.C., Petkova, E., Neria, Y., Van Meter, P.E., Zhao, Y., … Marshall, R.D. (2015). Is exposure necessary? A randomized controlled trial of Interpersonal Psychotherapy for PTSD. American Journal of Psychiatry, 172, 1-11.

Post-traumatic stress disorder (PTSD) is a condition caused by experiencing or witnessing a terrifying event. Symptoms may include flashbacks, nightmares and severe anxiety, as well as uncontrollable thoughts about the event. PTSD has a lifetime prevalence of 6.8%, which makes it a highly prevalent disorder. The main technique of empirically validated psychological treatments for PTSD involve exposing patients to safe reminders of the trauma including memories, with the intent of extinguishing the fear responses. This is the basis of cognitive behavioural therapy (CBT) with prolonged exposure, which is a consensus treatment for PTSD. However, not all patients benefit from CBT with prolonged exposure, and such treatment may be too difficult for some patients and therapists to tolerate. Markowitz and colleagues argued that PTSD symptoms reflect interpersonal issues including interpersonal withdrawal, mistrust, and hypervigilence. Interpersonal psychotherapy (IPT) is a time-limited efficacious treatment for depression that was adapted for this study for non-exposure based non-CBT treatment of PTSD. IPT was modified so that the first half of treatment focused on recognizing, naming, and expressing feelings in non-trauma related interpersonal situations. The second half of treatment focused on common IPT themes such as role disputes and role transitions. The authors argued that IPT helps individuals with PTSD gain mastery over social interactions and mobilize social supports. The authors conducted a randomized controlled trial that had a sufficient sample size to test a hypothesis of “non-inferiority”, that is to adequately test if PTSD and exposure based CBT were equally effective. Both treatments were compared to a progressive muscle relaxation (PMR) control condition. In all, 110 participants with chronic PTSD were recruited and randomized to IPT, CBT, or PMR. Most patients reported trauma of 14 years duration from either sexual or physical abuse, and half had a current comorbid depression. All three interventions resulted in large significant reductions in PTSD symptoms. IPT (63%) and CBT (47%) were not significantly different in rates of response (i.e., in which response was defined as 30% improvement in a clinician administered PTSD scale), but IPT had a significantly higher response rate than PMR (38%). Patients with comorbid depression were more likely to drop out of CBT with prolonged exposure than IPT.

Practice Implications

The results of the study suggest that IPT and CBT with exposure were equally effective in reducing symptoms of PTSD. It is important to keep in mind that this is one well-conducted trial that needs to be replicated by independent researchers in order to establish if the findings are truly reliable. Nevertheless, the findings contradict the widespread belief that patients with PTSD require exposure-based treatment in order to improve. IPT may be another option for the treatment of PTSD, especially for patients who cannot tolerate the prolonged exposure. Patients with comorbid depression may have the most difficulty tolerating prolonged exposure therapy, and so they may benefit from IPT as an alternative. IPT may help patients gain abilities in social interactions and social support, which may make it easier for them to spontaneously expose themselves to recollections of trauma.

Roest, A.M., de Jonge, P., Williames, G.D., de Vries, Y.A., Shoevers, R.A., & Turner, E.H. (2015). Reporting bias in clinical trials investigating second-generation antidepressants in the treatment of anxiety disorders: A report of 2 meta-analyses. JAMA Psychiatry, doi:10.1001/jamapsychiatry.2015.15.

Previous research has shown that the effects of antidepressant medications for treating depression may be over estimated by as much as 35%. This occurs because of publication bias, which refers to the tendency among researchers and editors to prefer to publish positive findings, and also occurs due to the occasional practice of the pharmaceutical industry to suppress negative findings. In these meta analyses, Roest and colleagues assess publication bias in the research of antidepressant medications to treat anxiety disorders (i.e., generalized anxiety disorder [GAD], panic disorder [PD), social anxiety disorder [SAD], post traumatic stress disorder [PTSD], and obsessive compulsive disorder [OCD]). Anxiety disorders are very common in the population, with an estimated year-prevalence of 12%. Second generation antidepressants (i.e., selective serotonin reuptake inhibitors and serotonin-norepinephrine reuptake inhibitors) are the primary pharmacologic treatment for anxiety disorders. Roest and colleagues were also interested in outcome reporting bias, which refers to mis-reporting non-positive findings as if they were positive findings, and spin which refers to interpreting non-positive results as beneficial findings. Positive findings refer to the pharmacological agent significantly outperforming a placebo, and non-positive findings refer to pharmacological agents not significantly outperforming a placebo. Pharmaceutical companies in the US must register any trial with the Food and Drug Administration (FDA) prior to starting the trial if the company wishes to apply for US marketing approval. And so, all medication trials and their findings, whether positive or non-positive must be listed in the FDA register. Despite being listed with the FDA, not all trials and findings get published in peer reviewed journals. This causes a problem for reviews and meta analyses that tend only to focus on published trials, and prescribing physicians tend only to read published trials and reviews. In their meta analyses Roest and colleagues compared findings from all FDA registered medication trials to those that were published in peer reviewed journal. Fifty seven trials were registered with the FDA but only 48 were published. Regarding publication bias, the proportion of studies with positive findings indicating efficacy of antidepressant medications in FDA trials was 72%, whereas the proportion of studies with positive findings in trials published in a journal was 96%. Overall, trials were 5 times more likely to be published if they were positive than if they were non-positive. Regarding outcome reporting bias, 3 of 16 trials that were non-positive in the FDA review were reported as positive in journal publications. Regarding spin, an additional 3 of the 16 non-positive trials interpreted non-positive results as if they were positive. Effect sizes in the FDA data was g = .33 indicating a small average effect size of the medications for anxiety disorders, but the effect size in published journals was g = .38 indicating a small to moderate effect. This represents a 15% over estimation of the effects of the antidepressant medications for anxiety disorders in the published literature.

Practice Implications

The effects of antidepressant medications for anxiety disorders appear to be over estimated by 15% in the published literature. This inflation is not as large as the 35% over estimation in the published literature of the effects of antidepressant medications for depression. By contrast, as I reported in a previous PPRNet Blog, publication bias in psychotherapy trials is small and has little impact on the overall estimate of psychotherapy’s efficacy. Effect sizes for psychological interventions for anxiety disorders are moderate to large, g = .73. Combining medications and psychotherapy only modestly improves efficacy of treatments, and medications may interfere with the efficacy of psychological interventions.

The Great Psychotherapy Debate: Since in April, 2015 I review parts of The Great Psychotherapy Debate (Wampold & Imel, 2015) in the PPRNet Blog. This is the second edition of a landmark, and sometimes controversial, book that surveys the evidence for what makes psychotherapy work. You can view parts of the book in Google Books.

Wampold, B.E. & Imel, Z.E. (2015). The great psychotherapy debate: The evidence for what makes psychotherapy work (2^nd edition). New York: Routledge.

In this part of the chapter on efficacy, Wampold and Imel provide convincing evidence from numerous reviews of meta analyses that the average effect size of psychotherapy across diverse treatments and patients is about d = .80. This is a reliable figure and is considered a “large” effect by commonly accepted standards. Put another way, the average psychotherapy patient is better off than 79% of untreated clients, psychotherapy accounts for 14% of the outcome variance, and for every 3 patients who receive psychotherapy, one will have a better outcome than had they not received psychotherapy. In other words, psychotherapy is remarkably efficacious. These effect size estimates are mostly drawn from randomized clinical trials that are highly controlled (i.e., therapists are highly trained and supervised, patients are sometimes selected to have no co-morbid problems, treatment fidelity to a manual is closely monitored, etc.). Some argue that the context of these trials renders them artificial, and that findings from these trials reveal little about psychotherapy practiced in the real world with complex patients. How do findings from controlled clinical trials compare to everyday clinical practice? Wampold and Imel review the evidence from three areas of research: clinical representativeness, benchmarking, and comparisons to treatment as usual. With regard to clinical representativeness, a meta analysis (k > 1,000 studies) coded the studies for type of treatment setting, therapist characteristics, referral sources, use of manuals, client heterogeneity, etc. The meta analysis found that therapies that were most representative of typical practice had similar effects to what is observed in highly controlled studies. With regard to benchmarking, a large study (N > 5,700 patients) compared treatment effects observed in naturalistic settings to clinical trial benchmarks. Benchmarks were defined as scores on an outcome (e.g., on a depression scale) that are within 10% of scores reported in clinical trial research. Treatment effects in naturalistic settings were equivalent to and sometimes better than those achieved using clinical trial benchmarks. Further, therapists in practice settings achieved the same outcomes in fewer sessions than in clinical trials. With regard to comparisons to treatment as usual, a meta analysis (k = 30 studies) for personality disorders looked at studies that compared evidence-based treatments tested in clinical trials to treatment as usual. The meta analysis found that evidence-based treatments were significantly more effective than treatment as usual with moderate effects. These results suggest that when it comes to personality disorders, special training and supervision, which are common in clinical trials, might be beneficial.

Practice Implications

Wampold and Imel argue that psychotherapy as tested in clinical trials is remarkably effective such that the average treated patient is better off than 79% of untreated controls. The evidence also suggests that psychotherapy practiced in clinical settings is effective and probably as effective as psychotherapy tested in controlled clinical trials. It is possible that therapists who treat those with personality disorders may benefit from additional training and supervision to improve patient outcomes in everyday practice.

Cuijpers, P., Koole, S.L., van Dijke, A., Roca, M., Li, J., & Reynolds, C.F. (2014). Psychotherapy for subclinical depression: A meta-analysis. British Journal of Psychiatry, 205, 268-274.

Subclinical depression refers to someone having relevant depressive symptoms but without meeting standard diagnostic criteria for a depressive disorder. Cuijpers and colleagues indicate that subclinical depression can be defined as meeting at least one but not more than four DSM core symptoms for depression. Subclinical depression is highly prevalent. About 50% of individuals with major depression have had a subclinical depressive disorder, and so subclinical depression may be a risk for developing major depression later on. Depression in general is associated with a high level of health and economic burden worldwide (see my June 2014 blog). Antidepressant medications are likely not more effective than a placebo in treating subclinical depression. Cuipers and colleagues examined whether psychotherapy is effective in treating subclinical depression, and whether psychotherapy reduces the subsequent occurrence of major depression. Cuijpers and colleagues report on a meta analysis of 18 studies of psychotherapy for subclinical depression representing 1,913 patients. Most of the studies were based on cognitive behavioral therapy (CBT). In order to compare the effects of psychotherapy for subclinical depression versus psychotherapy for major depression, they also included 56 studies of psychotherapy for major depression. Psychological treatments had a small to moderate effect on subclinical depression (g = .35) that was statistically significant. Psychotherapy significantly reduced the incidence of major depressive episodes by 39% at 6 months follow up, and by 26% at 12 months follow up. The effect of psychotherapy for major depression (g = .63) was significantly larger than the effect of psychotherapy for subclinical depression (g = .35). No differences were found between CBT and other forms of psychotherapy for subclinical depression.

Practice Implications

The results of this meta analysis indicate that psychological treatment of subclinical depression is moderately effective, and may reduce the incidence of major depression in the longer term for some. Effect sizes of psychotherapy for subclinical depression were likely underestimated because the type of control groups used in these studies affected study quality. However, even after controlling for study quality, the effects of psychotherapy for subclinical depression were still smaller than effects for psychotherapy of major depression. Although the number of studies comparing CBT to other therapies is small, the findings are similar to other meta analyses that indicate that several psychotherapies are effective treatment options for depressive symptoms.

Xie, C.L., Wang, X.D., Chen, J., Lin, H.Z., Chen, Y.H., Pan, J.L., & Wang, W.W. (2015). A systematic review and meta-analysis of cognitive behavioral and psychodynamic therapy for depression in Parkinson’s disease patients. Neurological Sciences, 1-11.

Parkinson’s disease (PD) is a neurodegenerative brain disorder that progresses slowly in most people. When dopamine producing cells in the brain are damaged or do not produce enough dopamine, motor symptoms of PD appear. Non-motor symptoms, including depression, apathy, and sleep disorders are also common so that in clinical settings about a 40% of patients with PD may have a depressive disorder. Depression is a top predictor of poor quality of life in patients with PD. Depression in PD is not well understood but may be due to neurobiological vulnerability and to psychological factors. Antidepressant medications are often prescribed for depression in PD but their efficacy is questionable. Xie and colleagues argue that long term use of some antidepressants may lead to worsening of some PD motor symptoms. In this meta analysis, Xie and colleagues examine the efficacy of brief psychological interventions, including cognitive behavioral therapy (CBT) and psychodynamic psychotherapy for depressive symptoms in PD. Twelve eligible studies were included in the meta analysis representing 766 patients with a mean age of 62 years (48% men). As an interesting note, 9 of the 12 studies were conducted in China and 3 were from the US or UK. Six of the studies used CBT for depression, and the remaining used psychodynamic therapy for depression in PD patients. Control conditions were often “treatment as usual”, and varied from antidepressant medication (e.g., Citalopram), nursing care, telephone calls, or no treatment for the depression. The effects of psychological interventions compared to control conditions on depressive symptoms were large, and remained large even after removing outlier studies. Outcomes for psychodynamic psychotherapy were better than for CBT, although both interventions resulted in large effects. There were also significant positive effects of brief psychotherapies on cognitive functioning, but not on quality of life. The authors were concerned that the quality of studies was variable and that many studies demonstrated a risk of bias. Further, most studies did not report outcomes at follow up periods.

Practice Implications

Significant depressive symptoms commonly occur in patients with Parkinson’s disease (PD). As a result, overall quality of life may be reduced in patients with PD. Medications for depression may be complicated by the neurodegenerative nature of PD – that is, effects of medications on depressive symptoms may be small and their neuro-motor side effects may be intolerable for some patients. This meta analysis by Xie and colleagues of 12 studies suggests that better research on psychotherapy for depression in PD needs to be conducted with adequate follow ups. Nevertheless, the findings suggest that brief psychological interventions may represent viable and effective alternatives for patients with PD who have a depressive disorder.