Dear Colleagues:
We are writing to you today to address some issues relevant to the
American Psychological Association’s PTSD Guidelines. As you are probably aware, there has been a backlash in response to these guidelines, including the circulation of an online petition (
https://www.thepetitionsite.com/takeaction/480/492/776/?cid=headerClick) asserting that the guidelines “did not take into account the evidence for intensive ‘talk therapy’ and for other treatments that have helped countless thousands of traumatized people get back on their feet and change their lives.” In this brief communication, we would like to give some perspective on this issue. Bear with us for a few paragraphs while we set the stage for some interesting evidence for supportive/psychoeducational treatment effects in PTSD.
As you may know, the guidelines were based on evidence from randomized clinical trials (RCT’s). RCT’s give an estimate of the strength of treatment effects when they are compared to either no treatment (e.g., a waitlist condition) or to some sort of “psychological placebo” condition. These “placebo” interventions are, in general, designed to control for the non-specific but nonetheless valuable common elements of treatment: things like establishing a supportive relationship with a caring clinician and receiving information about the disorder to normalize symptoms and to spur problem-solving efforts. Sometimes more complex treatments are also included as “placebo” conditions. This is the case with some PTSD treatment trials; for example, Present-Centered Therapy (PCT) has been used as a comparison condition to trauma-focused cognitive-behavior therapy. PCT is an interesting comparison condition because it is not a trauma-focused therapy and instead emphasizes active problem-solving training to minimize the effects of trauma and trauma-related responses on patients’ lives.
A strength of RCTs, then, is that a treatment undergoing evaluation must offer an outcome that is significantly stronger than the waitlist and/or active comparison conditions. Moreover, when it comes to summaries of studies as used in meta-analyses and the APA Guidelines process, a treatment must beat these comparison conditions repeatedly, so that we know that the overall outcomes averaged across trials represent more than just a lucky finding. Reliability of findings matters, because it means that benefit was observed across studies, across research settings, and, in the case of PTSD, across many different patients with many different traumas.
In the petition, there is a complaint that the guidelines were reliant on this sort of averaging process from well-controlled trials. Specifically, the petition asserts that, “Due to the exclusive reliance on RCT outcomes in formulating the guideline, other legitimate research findings crucially and commonly supporting additional forms of PTSD treatment were not considered in the assignment of ‘strong’ recommendations.” Ironically, however, the information provided by RCTs
can give us some additional insight into some of the alternative forms of treatment, insofar as they were included as control or comparison interventions in addition to the targeted treatment. As such, let’s turn our attention to the research evidence for
the strength of these comparison interventions.
This process is aided by the very recent publication of a meta-analysis by Carpenter and associates (2018) (
https://www.ncbi.nlm.nih.gov/pubmed/29451967) examining the efficacy of CBT relative to “psychological placebo” conditions rather than to waitlist control conditions. Accordingly, this is a meta-analysis of the ability of targeted treatments to beat a higher standard – namely to outperform clinical interventions emphasizing common elements like clinician support and psychoeducation. In general, the answer from this meta-analysis for anxiety disorders is that CBT offers efficacy over and above these alternative approaches. More importantly for the topic at hand, the Carpenter et al. (2018) meta-analysis can also be used to provide an index of the strength of these comparison treatments for PTSD. We recently asked the first author to break out this statistic from the overall results, and here is the result:
For all of the psychological treatment control conditions (e.g., supportive treatments and PCT) included in the Carpenter (2018) meta-analysis, the
within-group Hedges’ g is 0.54 (95% CI = 0.41-0.67). This includes 13 studies total.
This means that there is reliable evidence of improvement from pre-treatment to post-treatment in these comparison conditions. This effect is certainly part of what is observed in clinical practice, and is consistent with the claim in the petition of evidence “for other treatments that have helped countless thousands of traumatized people get back on their feet and change their lives.” But how strong are these treatment effects? After all, the within-group pre-treatment to post-treatment effect size does not control for non-treatment-related recovery effects, making it difficult to get an estimate of how well a comparison treatment works relative to no treatment. To estimate this effect we would need to include a waitlist condition in addition to the common elements comparison condition, which is rare to find in an RCT. Fortunately, we were able to identify four stalwart PTSD research teams that have done exactly this. Based only on these four studies, it turns out that “psychological placebo” treatments used in RCT’s have an advantage over no treatment (waitlist control) on the order of a moderate effect size: d = 0.57 (between-groups Cohen’s d effect size at post-treatment).
We have to admit some surprise at the strength of this effect for supportive and PCT interventions for PTSD. We tend to think of PTSD as difficult to treat, perhaps because it is often emotionally painful for clinician and patient alike. In this context, we are surprised at how well supportive, psychoeducational, and problem-solving treatments did. As such, we agree fully with one specific assertion from the authors of the petition – a range of treatments can offer benefit to PTSD patients!
But of course, the purpose of the Guidelines was not to show that patients can get some benefit from any treatment. The purpose, as we understand it, was to identify the treatments that have reliable evidence for
strong benefits. Based on our review of the APA PTSD Guideline effect sizes, treatments emphasizing prolonged exposure (d = 1.27) and treatments emphasizing other cognitive-behavioral and cognitive interventions (d = ~1.0) have effects that are roughly
double that of the comparison treatment estimate noted above (d = .57). That is, to our eyes, the guidelines correctly direct attention to the treatments that seem to be offering the very best outcome to patients as evaluated by repeated controlled trials. Other treatments provide benefit, but the most reliable evidence for the strongest benefits is for trauma-focused treatments emphasizing prolonged exposure, cognitive behavioral therapy, cognitive processing therapy, or cognitive therapy. It is for these reasons and others that clinical researchers are now circulating a very different petition (
https://www.thepetitionsite.com/780/537/970/support-the-apa-ptsd-treatment-guidelines/) to underscore their support for the PTSD Guidelines.
A core feature of the mission of Division 12 of APA is “to represent the field of Clinical Psychology through encouragement and support of the integration of clinical psychological science and practice in education, research, application, advocacy and public policy…” As clinical researchers and clinicians we think the guideline process is valuable for encouraging treatment approaches that have shown their mettle under the bright lights of scrutiny that controlled trials offer. Given these successes, it seems only right that the APA guidelines encourage us to shine our own attention on these treatments when we consider what is right for our patients suffering from PTSD. This is what guidelines are supposed to do.
Sincerely,
Michael W. Otto, PhD
Eugenia Gorlin, PhD
Hayley E. Fitzgerald, BA
Department of Psychological and Brain Sciences, Boston University
