by Michael D. Anestis, M.S.
In the most recent issue of the prestigious Journal of Consulting and Clinical Psychology, Rebecca Stewart and Dianne Chambless of the University of Pennsylvania published a meta-analysis that examined a pivotal question regarding the movement for empirically supported treatments (ESTs): does cognitive behavioral therapy (CBT) work for anxiety disorders in "real world" settings as well as it does in controlled laboratory settings? This question is pivotal because it provides insight into whether the work conducted by researchers translates into productive results for clinicians who treat clients in less controlled settings.
We have discussed meta-analyses and effectiveness trials in prior articles; however, before discussing the results of this particular study, I would like to provide a quick definitions for each term for readers who are new to the site. Remember, these are quick, simplistic definitions, so for a better understanding of either concept, I would suggest further reading on the matter.
Meta-analysis - In most studies, researchers collect data from a large group of individuals and then examine that data to see if their hypotheses are supported. Some researchers look to see if one group improves significantly more than another group. Other researchers look to see whether one variable is related to another such that, as one increases the other also increases or vice versa. In a meta-analysis, researchers look at a group of studies rather than a group of individuals to see if a particular relationship is found consistently by different researchers and, if so, how substantial that relationship might be.
Effectiveness study - In an efficacy study, researchers often use very strict criteria for who can take part and receive treatment. This might mean that participants can not meet criteria for more than one diagnosis, can not have a history of prior suicide attempts, or can not have any prior history of psychotherapy. Additionally, therapists in these trials are typically highly trained in the relevant therapy, use a manual to guide treatment, and are strictly supervised to ensure that the therapy is administered in a manner consistent with the protocol. The researchers do this in order to ensure that whatever results they find are best accounted for by the therapy being administered; however, such procedures often raise concerns as to whether or not the results will be representative of what happens in "real world" clinical settings. Effectiveness studies, on the other hand, do not use these exclusion criteria or demands on supervision. They simply examine the degree to which a particular therapy produces results in a naturalistic, "real world" setting.
So, in the study I will discuss today, Stewart and Chambless (2009) examined a group of studies (meta-analysis) to see if CBT produced strong results in the treatment of anxiety disorders in "real world" clinical settings (effectiveness studies). In this particular meta-analysis, the authors looked at treatment outcomes for individuals who met criteria for panic disorder (PD), social anxiety disorder (SAD), post-traumatic stress disorder (PTSD), generalized anxiety disorder (GAD), and obsessive compulsive disorder (OCD). The results were as follows:
Effectiveness in "real world" settings:
The authors found large effect sizes for CBT in reducing disorder-specific symptoms for every condition examined in the meta-analysis. In other words, OCD symptoms substantially diminished for individuals seeking treatment for OCD, PTSD symptoms substantially diminished for individuals seeking treatment for PTSD, and so on. The large effect size is important, as this means that CBT produced clinically significant improvements reflective of a dramatic increase in participants' quality of life. Additionally, CBT resulted in substantially reduced symptoms of depression for participants regardless of which anxiety disorder(s) for which they met criteria.
Effectiveness relative to control groups:
Not every study examined in this meta-analysis included a control group, but Stewart and Chambless (2009) examined this question for those that did. Some control groups were comprised of individuals on a waitlist awaiting treatment. Treatment conditions are compared to waitlist conditions to ensure that the mere passage of time does not better account for symptom reductions. Other control groups were comprised of participants receiving "treatment as usual." In other words, in certain studies, CBT was compared to less structured, often eclectic approaches to treatment. The authors found that CBT produced significantly better results than did control groups, again with a large effect size. Using a binomial effect size display, the authors found that the results translated to a 78% improvement for individuals receiving CBT as compared to a 22% improvement for individuals in the control conditions.
Comparing effectiveness results to efficacy results:
Even with strong results in effectiveness studies, the question remains to what degree these results are as strong as those found in efficacy studies. The authors reported that the effect sizes in the effectiveness studies were, in large part, consistent with those found in efficacy studies. In other words, moving CBT out of the lab and into "real world" practice does not appear to diminish its impact.
Does the utility of CBT depend upon the "clinical representativeness" of the sample?
Another potential criticism of effectiveness studies might be that it is possible that stronger results are only found when the participants are not similar to those found in "real world" practice. The results in this study did not support that fear. Although an extremely small relationship was found between clinical representativeness and effect sizes, this relationship disappeared when measures with extremely low reliability were removed. Perhaps more interestingly, the authors found that the results improved when medication was allowed and when participants were not randomized to treatments (e.g., CBT versus treatment as usual). Additionally, they found that results were less impressive when the therapists were not trained in CBT, when therapists were not asked to follow a manual, and when therapists received little or no monitoring to ensure that they are implementing treatment according to the prescribed protocol. These effects were all small; however, they point towards the utility of treatment manuals and legitimate training in "real world" practice.
So, overall, what do these results tell us? CBT appears to not only be an efficacious treatment for anxiety disorders, but also an effective one. In other words, the impact of CBT is not due to the machinations of researchers disconnected from true clinical practice who control the types of clients they treat and only measure variables that will support their claims. Instead, the impact of CBT in laboratories is echoed by equally compelling findings in real world practice.
Importantly, CBT is not the only empirically supported treatment. Remember, Division 12 of the American Psychological Association has comprised a growing list of various treatments for particular mental illnesses. CBT is simply researched more often than most other forms of treatment. Any researcher or clinician who believes in alternative approaches is free to test their hypotheses and share them with the world. Nobody is stopping them from doing so.
If you or somebody you know is struggling with anxiety, please consult our EST clinics page to find local effective care. If you do not see any local clinics listed, try calling the nearest university clinic and asking them where you might be able to receive empirically supported treatments. If they do not know what you mean, direct them to this site and tell them they are fee to contact us to discuss the issue.
If you would like to learn more about the effective treatment of anxiety disorders, we recommend the following resources as well as several others that can be found in our online store:
- Feeling Good: The New Mood Therapy Revised and Updated
- David Burns, M.D.
- When Panic Attacks: The New, Drug-Free Anxiety Therapy That Can Change Your Life
- David Burns, M.D.
- Mastery of Your Anxiety and Panic: Workbook
- David Barlow, Ph.D.
- Stop Obsessing!: How to Overcome Your Obsessions and Compulsions
- Edna Foa, Ph.D.
- Prolonged Exposure Therapy for PTSD: Emotional Processing of Traumatic Experiences Therapist Guide - Edna Foa, Ph.D.
- The Worry Cure: Seven Steps to Stop Worry from Stopping You
- Robert Leahy, Ph.D.
- Anxiety Free: Unravel Your Fears Before They Unravel You
- Robert Leahy, Ph.D.
- Treatment Plans and Interventions for Depression and Anxiety Disorders
-Robert Leahy, Ph.D.
Mike Anestis is a doctoral candidate in the clinical psychology department at Florida State University
Subscribe in a reader
