EMDR FORUM ARCHIVE, 2000
Exposure and retraumatization
Cahill · 02/05/03 at 9:48 PM ET
Dear client: Thanks for the question. Several people have voiced the exact same concern. I think that there two aspects about your question that make it difficult to provide a straightforward answer. The first issue is a semantic one: What do we mean by the use of the terms "trauma" and "retramatization." When I use the terms, I try to carefully stick to the DSM-IV definintion of a trauma which involves to requirements: The first requirement is objective exposure to a situation that threatens the life or physical integrity of self or other. Thus, being physically assaulted, witnessing a murder, etc. are potentially traumatic events. Being told by a significant other that they no longer love you and want to end the relationship (just as an example) may be terribly upsetting but, by the DSM definition is not a trauma. The second requirment a subjective reaction to the event involving intense fear, terror, horror, or helplessness. A person may be confronted with a life/death situation, but if it didn't have an emotional impact, then it too is not classified as a trauma. Now, if we apply this definition to the issue of "retraumatization" in exposure therapy, we have to ask the question is exposure therapy a potentially traumatic event? I would argue that it is not. Talking about a past trauma, although upsetting, does not involve real threat to the life or physical integrity of self- or other. To give a personal example, I cut my thumb on a table saw when I was in High School. I was physically injured and the injury certainly could have been worse. I was lucky in that no bone, nerves, or tendons were damaged. I just lost a little meat. At one time, imagining the saw cutting into my skin or hearing the sound of a saw hit a knot in a piece of wood used to cause me to feel anxious. It was quite unpleasant. However, the thought of getting cut did not cause physical injury. So, from this perspective, exposure therapy may elicit unpleasant thoughts and feelings, but there is no real threat of harm and therefore is not a trauma. If exposure therapy is not a trauma, then it cannot be traumatizing, much less "retraumatizing." Now, you are free to agree or disagree with the definition, but then please specify what you mean when you use the term trauma. 2. Given the definitional issues discussed in number 1 above, we need to recast your question, which is perhaps better phrased in terms of whether exposure therapy puts people at risk for increases in their PTSD symptoms and/or whether there are people for whom exposure therapy seems to have little or no therapeutic effect? Moreover, if either of these are true, is it to a greater or lesser extent than occurs in other forms of treatment? Although there is much speculation on this issue, and lots of opinion based on clinical experience, there is not much systematic data. However, I'll try to summarize what we do and do not yet know. Second, exposure therapy has been compared with several other treatments for PTSD including relaxation, EMDR, anxiety management (called stress inoculation training), and cognitive therapy. If exposure therapy were causing a significantly greater number of instances of patients gettting worse or not getting better than these other treatments, you would expect that the overall outcome would be that the other treatments should wind up with superior outcomes. While the specific results may shift around a bit from one study to the next, the overall pattern is that all these treatments seem to result in very similar outcomes. Third, researchers (including our own group) are starting to be more systematic in investigating this issue. There are a few studies on point: Tarrier et al. (1999) compared imaginal exposure therapy for PTSD with classic cognitive therapy. On avearge, both groups produced similar outcomes. However, on one measure, the CAPS, the exposure therapy group had a greater amount of variablity than in the cognitive therapy group. One possible explanation for this is that there are two subgroups within the exposure therapy group: one that got a whole lot better, and a second that didnt' do so well. When you average these two groups, it looks everyone is doing OK. They then did something very interesting. They classified people as either having gotten worse on the CAPS by at least one point or not. When they did this, they found that 30% of the exposure therpay subjects "got worse" compared to only 10% in the cognitive therapy group. I put "got worse" in quotations because their criteria of a single point worse could simply be achieved through measurement error, and therefore it is not clear whether these people really got worse, or simply didn't get better. Putting the semantic issue aside, the results indicated that more people benefitted (or fewer people got worse) from cognitive therapy than for exposure therapy. This is a finding of concern. But we need to know at least two more things. First, how many people would have "gotten worse" (or not gotten better) had they not received treatment? Tarrier et al. did not inlcude a waitlist group so we don't know if the 30% in exposure therapy is any more or less than would have occurred without treatment. So maybe exposure therapy caused more people to "get worse" (not get better) than would have occurred with the passage of time OR maybe exposure therapy was less effective than cognitive therapy in reducing the number of people who "get worse" (not get bettter). The second thing we need to know is whether the Tarreir results are representative of other studies or if they are a single finding. There is now a published study of a treatment that combined affect regulation skills training with exposure therapy for the treatment of PTSD in adult women who were victims of childhood abuse (Cloitre et al., 2002). They conducted a similar analysis and found that 25% of participants in the waitlist group got worse, compared to only 5% in the treatment group. So, we now know that at least some people even "get worse" in the absence of treatment, and that treatment actually reduces the percent of people who "get worse." Thus witholding treatment is probably as dangerous as administering treatment, and administering treatment is more likely to result in improvement than witholding treatment. I have used some of our own data to look at this question in our studies and found that about 10% of people on waitlist get worse by at least on point on our measure of PTSD symptom severity, compared to between 0-5% in our various treatment conditions, including exposure therapy (alone or with cognitive therapy added) and stress inoculation training (alone or combined with exposure therapy). Importantly, exposure therapy was not worse than stress inoculation training. So, our results are consistent with those of Cloite et al. We don't yet know why our results differ from Tarrier et al., but it is certainly the case that exposure therapy can be done safely. Some of my colleagues recently published an interesting set of analyses that looked at people getting worse early in therapy (the above studies looked at pre- vs. post-treatment). Foa, Zoellner, Feeny, Hembree, and Alvarez-Conrad (2002) did find that a minority of participants did show an initial increase in the PTSD symptoms right after imaginal exposure was started. Moreover, this increase was operationally defined so that it had to be larger than would be expected by measurement error alone (one of the limitations of the Tarrier et al. study). That's the bad news. The good news is that the increase was only temporary and it did not prevent people from completing treatment. Moreover, both the people who showed the intial symptom worsening and the people who didn't show this pattern both had good overall outcomes at the end of treatment. Nishith et al. (2002) compared exposure therapy with cognitive processing therapy and looked at the course of symptom change over treatment. They found that both treatments were associated with a recovery curve that was best fit by a quadratic function, meaning that in BOTH therapies, there was a slight increase in symptoms in the early therapy sessions before switching over to a dramatic decrease. So symtpom worsening does occur in exposure therapy, but also in other treatments, and the symptom worsening is temporary. No one has looked at this issue with EMDR or stress inoculation training so we can't say if it is the same of different. Ironson et al. (2002) also compared EMDR with exposure therapy, and reported SUDs at the beginning and end of the first active treatment session session. The results produced an interesting pattern. SUDs were higher at the beginning of treatment for EMDR (79) than the were for exposure therapy (55). By the end of the session, things were the opposite. SUDS were lowere in EMDR (32) than exposure therapy (61). I'm not sure what this means, except that EMDR can be more "stressful" or less stressful than exposure therapy (and vice versa) depending on when you take the measurement. Client, I hope that you're still with me at this point. If so, I'll try and sum things up. 1. Some people with PTSD get worse if you withold treatment, a few get better, and most show no real changge. 2. Some people with PTSD may get worse or not improve if you treat them with exposure therapy. However, more people get better and fewer people either stay the same or get worse when you treat them with exposure therapy than if you withold treatment. 3. We have little information about whether treatment with exposure therapy produces more, less, or the same number of cases that get worse/don't change than other treatments. The extant data are mixed. My own best guess is that at the end of the day, we will find few differences between treatments on this measure, just as we've repeated failed to find convincing evidence for the superiority/inferiority of one form of cognitive-behavior treatment with another. 4. Symptoms worsening can occur in exposure therapy and cognitive processing therapy, but only a minority of patients show this pattern and it is temporary. It most likely (in my opinion) reflects accessing the trauma memory. If so, then we'd expect a similar phenomenon in EMDR. This issue has to be empirically studied for EMDR. 5. There really is no convincing evidence that exposure therapy is more stressful than EMDR. So, I value your concern about the safety of exposure therapy. I'd further say that this issue is not unique to exposure therapy, but is relevant to EMDR and cognitive therapy as well. It is also of concern to witholding treatment. At present, fear of symptom worsening, however, is not a reason to withold treatment.
First, there are quite a number of studies on exposure therapy for PTSD among a range of different participant populations. Of these studies, I can only think of two studies (Boudewyns et al. and Pitman et al. 1996, both studying Veterans) in which there appeared to be little or no positive effect of exposure therapy. In all the other studies, exposure therapy was associated with overall positive outcome. Also, it should be noted that Boudewyns and Pitman have also published EMDR studies in which treatment benefit was similarly limited. So, the failure to observe an overall treatment effect with veterans is not unique to exposure therapy, and other studies of exposure therapy (and EMDR) with veterans have yielded positive results.
Two studies have compared exposure therapy and EMDR in terms of how stressful patients found the treatments. Devilly & Spence (1999) asked patient to fill out a questionnaire at the end of treatment to rate how distressing the found EMDR or a combination of exposure therapy plus stress inoculation training plus cognitive therapy. Now, there is some controversy about this study. The study has been criticized for alterning the EMDR procedure. I personally don't find this critique to convincing. Rather, I have critcized the study for not having used complete random assignment and for switching the assigment of the last patient from EMDR to exposure therapy on the basis of having conducted a preliminary analysis of the data which found exposure therapy to be superior to EMDR. They defend their actions on ethical grounds, but I don't agree. Be that as it may, patients used a scale to rate treatment stress that ranged from 8 to 72. Both groups had an average distress score of 38, no difference at all. Thus, this study provides no evidence that exposure therapy is any more stressful than EMDR.
Replies:
|
| Behavior OnLine Home Page | Disclaimer |
Copyright © 1996-2004 Behavior OnLine, Inc. All rights reserved.
