Psychotherapy involves talking through emotional difficulties with a trained therapist and is often what springs to mind for many people when psychological treatments are mentioned. But being well-known and being effective are very different things.
Significant and positive benefits were found in 80% of instances where psychotherapy was used as a treatment, according to Evangelou and colleagues of the University of Ioannina, Greece.
This stood after reviewing 247 unique meta-analyses involving data from over 5000 randomly controlled trials. These meta-analyses included studies where any form of psychotherapy aimed at any mental health condition was used. The only condition was that a minimum of ten studies were taken into account.
At first impressions, seeing improvements in 80% of people receiving psychotherapy sounded impressive. But after a closer look at the strength of the evidence, only 16 of the 247 meta-analyses were found to be compelling.
There were other issues identified by Evangelou and colleagues. A current debate about how acceptable it is to compare studies with such different content could be applied to the meta-analyses here.
Also, there were issues of biases. The researchers found that many of the studies which were smaller reported larger effects. The larger a sample size, the more generalizable the results, suggesting caution should be taken with results of smaller studies.
In addition, the researchers thought it was very likely that many studies with negative findings were not published, giving readers an unrealistic, imbalanced view of the effectiveness of psychotherapy. This is something that could easily be overcome, simply by pre-registering studies and the analysis that will be undertaken before the research takes place.
It may well be that it is the way in which psychotherapy is assessed in terms of effectiveness and how research in this area is published that need revising. Given the subjective nature of psychotherapy itself, is this feasible?
For a more in-depth look at this issue and others, see the BPS Research Digest page here: https://digest.bps.org.uk, edited by Christian Jarrett.