Treatment for Anxiety
I wrote this reflection for my problem identification class at the University of Rochester. The professor asked us to reflect on a particular reading, the reading for this reflection was Bandelow's (2015) Efficacy of treatments for anxiety disorders: a meta-analysis. I have included my reflection below.
Treatment for Anxiety
This week, I chose to reflect on Bandelow's (2015) Efficacy of treatments for anxiety disorders: a meta-analysis. My initial reaction upon reading this piece was that I walked into a grudge match! Bandelow opens the paper with a very striking abstract, claiming his paper is unique, comparing all major forms of anxiety with treatment efficacy across drug-based treatment (pharmacological), therapy-based treatment (psychological), and controls, such as placebo pills or waitlists. Bandelow cites his usage of meta-analysis, displaying largess in numbers, touting 234 studies analyzed covering 37,333 patients. Following this, Bandelow slams home his finding that medications were associated with significantly better patient outcomes, an effect size of 2.02, compared to therapy, with an effect size of 1.22. Bandelow does caveat that in some cases, CBT performed better than placebo, but he closes his abstract with the hard-hitting remark that, essentially, psychotherapy is no different from placebo.
In the introduction of Bandelow's paper, he comments on the National Institute for Health and Clinical Excellence (NICE) and how their views seem to differ from the results he has collated in this paper. It appears to me that this paper is a hit job against "NICE." I also had a chuckle when I saw at the very end of the paper, under "Conflict of Interests," that Bandelow and his co-author Wedekind are on boards (speaker boards or advisory boards) for Meiji-Seika, Lundbeck, Pfizer, Servier, AstraZeneca and Essex Pharma, all pharmaceutical companies promoting the drugs that Brandelow's report says are superior to psychotherapy. As we delve further, Bandelow obfuscates the details of his approach with technical statistical language. The parts that remain easy to understand for me are the parts that support the pharmaceutical companies' desires to promote drug use and taint psychotherapy as no better than popping a sugar pill.
As I read, I continued to wonder how 234 studies, all using different controls, variables, assessments, samples, pools, measurements, and time frames, could be consolidated into such a succinct headline statement that drugs work and that therapy is no better than placebo. In any case, a couple of interesting tidbits I extracted from the paper include some more knowledge about placebos; I did not realize that pill placebo has changed in effectiveness over time, "from 0.87 in the years 1983–1992 to 1.71 for the years 2003–2013", this was very interesting to me and made me think about the social part of biopsychosocial, how societal belief on pills has dramatically increased their placebo effect. I was also very interested in the section entitled "Active versus control effect sizes." In this section, the authors explain that while their overall meta-analysis shows drugs as superior, these differences go away when they look at direct or head-to-head comparisons between smaller sets of reports. We get varied results, especially with CBT, which shows a superior effect to control in these scenarios, as do relaxation and mindfulness-based techniques, with mixed results.
I was also interested in how Bandelow attempted to preemptively defend his own potential bias in allegiance with the pharmaceutical companies by doing statistical analysis on "allegiance effects," where he attempts to show that even if he and other authors of pharma-based studies do have allegiance, it is not significant in the results. He also then tries to pre-emptively tackle the objection that therapy-based procedures need a longer time to take effect compared to drugs, showing that the mean research period for drugs was ten weeks. In contrast, for therapy, it was 12 weeks, and even 21 weeks for psychotherapy. Bandelow also notes that it could be hard to say because many patients studied in the therapy research could also be on medications. He also notes that patients in drug studies may have had more severe anxiety than those in therapy studies.
I was shocked overall by this paper to see what I thought to be a blatant manipulation of statistics and language in order to trick readers into taking more drugs and avoiding psychotherapies. I kept thinking we could not draw such conclusions when Bandelow introduced so many confounding variables in the meta-analysis of 234 studies over a 20-year period. Bandelow had an average of 10 weeks of duration used in the drug research; we do not know if the drugs maintained efficacy over extended periods, especially over the decades and lifetimes people seem to be using them. Bandelow only superficially touched on side effects; if a patient is on drugs for years (or decades) that were tested only on a 10-week sprint, it can create horrible unforeseen consequences, not to mention potentially block the resolution of the underlying trauma that caused the anxiety condition in the first place.
Furthermore, we do not know the duration of placebo control effects; one could imagine that they only last a shorter time and disappear once usage ceases or the novelty of a placebo wanes. Suppose placebos were studied longitudinally versus psychotherapy, without the use of drugs. In that case, we possibly could see significant results in psychotherapy curing the underlying anxiety condition rather than placing it at bay. My personal experience has been this: I have been meditating 40 minutes a day for five years, I have seen a therapist weekly for about ten years, I sleep well, exercise and eat well, I reduce stress, and my anxiety has been effectively cured; I take no medication. I plan to consider all of this when working with my clients. After reading this piece, I am grateful to be better informed and equipped about the criticisms volleyed at therapy as an intervention so I may strengthen my ability to deliver well-being in my encounters.
