According to a follow-up study, the prior indicating an equivalent publication bias in scientific researches reporting the efficiency of antidepressant drugs, scientific literature does not portray psychological treatments for depression correctly – the latter are not as rosy as they seem to be.
Psychological Treatments Aren’t All They’re Hyped Up To Be: What’s The Problem?
Co-author Steven Hollon, Gertrude Conaway Vanderbilt Professor of Psychology and his colleagues from Oregon Health & Science University, VU University Amsterdam and the University of Groningen state that the results don’t imply that psychotherapy has not positive effects. Psychotherapy does work, but not as well as it is portrayed in scientific literature.
The main dilemma occurs because of clinical studies – studies demonstrating the positive outcomes of depression are preferred and published more readily than studies which show less favorable results.
“It’s similar to flipping many coins and keeping only those which come up heads”, explained Hollon.
Conducting An Analysis
The team of researchers identified all the US National Institutes of Health grants that funded clinical trials involving psychological treatments for depression between 1972 and 2008. Upon analysis, it was seen that about a quarter of the grants – 13 out of 55 – had not published the results of their trials.
The researchers of these 13 unpublished studies were contacted and requested for their results. Analyzing the published and unpublished data together, a series of meta-analysis concluded that psychotherapy is effective, but its efficacy is usually inflated and over-stated via publication bias.
Highlights And Suggestions
“These results highlight the aspect of publication bias that occurs in psychotherapy, corroborating what has been seen previously with antidepressants and various medication”, stated co-author Erick Turner, Associate Professor of Psychiatry and Pharmacology at OHSU School of Medicine, director of the previous study conducted in 2008 with respect to antidepressants.
“Articles published in journals are generally peer reviewed. However, this process has loopholes; allowing the benefits of treatments to be overstated and possible harms to be understated”, explained Turner. “This skewed information affects health care providers and, ultimately, their patients”.
A way to create more transparency is for journals and funding agencies to archive the original proposals along with raw data from the clinical trials, both published and unpublished. This would allow reporting bias to be detected and amended. The study can be found in the journal PLOS ONE.