Does Publication Bias Overstate the Effectiveness of Psychotherapy?

Depression is one of the most common mental health disorders in the US, according to the National Institute of Mental Health.   And most of us have known a loved one, friend or acquaintance impacted by the disease.   But depression can still be misunderstood and misdiagnosed, or in some cases have no treatment sought at all.

To be classified as Depression, The National Institute of Mental Health identifies it as a period of two weeks or longer during which there is either depressed mood or loss of interest or pleasure, and at least four other symptoms that reflect a change in functioning, such as problems with sleep, eating, energy, concentration, and self-image (NIMH, 2015).

As recent as 2013 the NIMH reported an estimated 15.7 million adults aged 18 or older in the U.S. had at least one major depressive episode in the past year – or 6.7% of all U.S. adults  (NIMH, 2013).  Currently, major depression accounts for:

  • 7 percent of all U.S. disability-adjusted life years
  • 3 percent of all U.S. years lived with disability

Depression is projected to become the 2nd leading cause of ill health by 2020.  Given the global impact of the disease, with 350 million people affected worldwide, effective treatment of Depression is critical.   According to the American Psychological Association, depression is very treatable, with different types of psychotherapy that are helpful for recovery in addition to the use of medications when necessary (APA, 2010).

But, a new study published in the journal Plus One may call into question the degree to which talk therapy is effective in relieving depression.  The study reviewed 55 National Institutes of Health grants awarded between 1972 and 2008, and found the results of nearly 25% were never published.  When researchers obtained the unpublished results, they brought down the overall efficacy of psychotherapy by about 25% (Hamilton, 2015).  Implying a publication bias in reporting study results.

Researchers acknowledge bias is a problem to be addressed in all fields of study.  But considered some changes that can be made to the grant process that require all results be published, not a select few.