Abstract
An impartial review of meta-analysis on the effectiveness of psychotherapeutic interventions has been published.
Note
This article was exposed to the machine translation from Ukrainian.
Background
Question about the effectiveness of psychological interventions have been occurring, perhaps, even from the beginning of their use for therapeutic purposes, from the time of Wilhelm Wundt and Sigmund Freud. Therefore it is not surprising that on this occasion it was published many scientific papers. To unify often disparate data frequently is used a special type of study - a meta-analysis, in which the results of all the available scientific work are reduced to a common denominator. The results of these studies are normally reflected in clinical guidelines.
One difficulty in interpreting randomized controlled trials are their possible bias, for example, they can be conducted by scientists who are interested in the results, because of support of one or another treatment method. Therefore, meta-analysis authors and organizations, creators of clinical guidelines, are faced against an important task of selecting scientific papers with highest quality, low probability of systematic errors, sufficient sample, adequate blinding of participants, etc.
It should be noted that according to the principles of evidence-based medicine, the appointment of a treatment method should be based on the existing evidence, taken from the above studies. And psychotherapy - is not an exception.
A team of scientists from academic institutions in Greece, the UK and Sweden, led by E. Dragioti from Medical School at the University of Ioannina, conducted an interesting study on the effectiveness of psychotherapeutic methods 1. Work was designed as a review of meta-analytical studies, another name - "umbrella meta-analysis". Authors paid much attention to the possible bias in already published meta-analyses.
Does psychotherapy work?
The researchers reviewed 37,662 articles on this subject, that have been presented in various electronic databases. Of these, 247 unique meta-analyses were selected, that pooled the results of 5157 randomized controlled trials.
The first what scientists have noticed is that overall 80% of all work showed a significant advantage of psychotherapy versus control treatment (placebo / other active therapy), which has to prove its efficacy (p <0,05). However, on closer inspection was spotted a great heterogeneity in these studies.
For example, one method could be extremely effective in one study, and in another the same method shows mediocre results. Heterogeneity is one of the indicators of possible bias. In addition, the researchers found that the effectiveness of psychotherapy is often exaggerated.
Overall, only 16 of 247 meta-analyzes cited robust evidence about the effectiveness of the investigational psychotherapeutic method. Strong evidence was found in the meta-analyses that examined the effectiveness of:
- Meditation therapy and cognitive behavioral therapy (CBT) in the treatment of anxiety disorders;
- Combination of different psychotherapeutical methods and CBT in depressive disorder;
- CBT in distress due to tinnitus;
- Various psychotherapeutic methods in irritable bowel syndrome (mental symptoms and daily functioning);
- Cognitive remediation to improve cognitive functioning in schizophrenia;
- CBT and psychological counseling for smoking cessation;
- Computerized CBT for depression and insomnia.
Thus, the researchers concluded that, indeed, psychotherapy works, but only a limited number of methods for conditions listed above. At least, if based on the evidence from these meta-analyses.
In what extent is psychotherapy effective?
As part of the analysis researchers calculated effect size for different psychotherapeutic methods. Note that the effect size <0.20 is considered as very small, 0.20-0.49 - small 0.5-0.79 - medium, 0.80-1.19 - large and > 1.20 - huge. Usually, if the effect size is at least medium, then it is a good indicator of effectiveness in comparison with the control method (placebo, another treatment, etc.). The effect size can be more or less than zero, which depends on compared values . The results are presented in Table 1.
Condition | Number of subjects | Type of psychotherapy | Control | Effect size |
---|---|---|---|---|
Anxiety disorders | ||||
Decrease in anxiety | 2,181 | Meditation | Other treatment or a waiting list | -0.52 |
Remission in children | 1434 | CBT | Waiting list | 0.22 |
Anxiety symptoms | 1369 | CBT | Placebo | 0.73 |
Major Depressive Disorder | ||||
Remission or a clinically significant response (6-8 weeks) | 1189 | Various types of psychotherapy + pharmacotherapy | Pharmacotherapy alone | 1.75 |
Social functioning | 1666 | Various methods | Treatment as usual or waiting list | 0.38 |
Symptoms of depression | 2313 | Various methods | Treatment as usual or placebo | 0.22 |
Dysfunctional beliefs | 1371 | CPT | Treatment as usual / waiting list / placebo | 0.51 |
quality of life | 2647 | Various methods | Treatment as usual / waiting list / placebo | 0.33 |
Clinical response in long-term | 1256 | Various methods + pharmacotherapy | Only pharmacotherapy | 2.93 |
Other conditions | ||||
Distress in tinnitus | 1770 | CBT | Treatment as usual / waiting list / other treatments | 0.62 |
Mental problem in IBS | 1656 | Various methods | Active therapy or waiting list | 0.41 |
Daily functioning in IBS | 1355 | Various methods | Active therapy or waiting list | 0.43 |
Cognitive functioning in schizophrenia | 1229 | Cognitive remediation | Treatment as usual | 0.42 |
Smoking cessation | 4375 | CAT | Self-help | 1.95 |
Abstinence from smoking in long time | 8223 | Psychological counseling | Advice of a doctor by the phone | 1.52 |
Symptoms of depression in insomnia | 1852 | Computerized CBT | waiting list | -0.39 |
Note remarkably high efficiency of psychotherapy for depression (CBT, interpersonal therapy, behavioral activation and family therapy). Concomitant use of psychotherapy and drug therapy can significantly improve the success of treatment. But the use of only psychotherapy or pharmacotherapy was associated with a significantly poorer response.