Always drink coffee

Hi Barbara:

I think perhaps I didn't quite express myself quite clearly on one point. I was not saying that the effects were especially strong in the as yet unpublished studies, compared to the currently published ones. What I meanst was that these currently unpublished studies (Rothbaum et al., Taylor et al.) further demonstrate (i.e., replication) the efficacy of exposure therapy. Also, both of these studies will be important contributions because they constitute direct comparisons between exposure therapy and EMDR.

We can say that exposure therapy and EMDR are effective treatments for PTSD. There is less data available on stress inculation training (also called anxiety management) and cognitive therapy, but the data we do have would suggest these treatments also work. So the good news here is that practitioners and patients have some choices.

Yes, the differential effects among these four therapies has been much smaller than their effects compared to waitlist, which is not surprising. If you compare to anti-depressent medications with a placebo, both meds will probably do better than placebo, but the two meds will probably perform similarly.

You raise an excellent question about whether the small differences among treatments may reflect the biases of the experimenters. This is ALWAYS a concern. Researchers attempt to take steps to reduce this. For example, Barbara Rothbaum, who took part in the first study of exposure therapy for rape victims (Foa et al., 1991), recieved official training in EMDR. In addition, they devised a treatment manual that met with Shapiro's approval. Finally, therapy sessions were video taped and a person that was selected by Shapiro reviewed a sample of the EMDR tapes to inspect the quality of the treament. Similar procedures were applied to the exposure therapy, except that the tapes were reviewed by people designated by Dr. Foa. The same thing goes for the Taylor et al. study. In fact, Louise Maxfield, who has posted on this forum and who has published on EMDR, was invovled in the running of that study. I defer to Louise if she has any specific comments she'd want to make about quality assurance procedures in that study. In short, alliance effects are always a potential confound. Good researchers take various measures to try and minimize it and to measure it so that it can be takin into consideration.

Finally, you ask how well do these treatments work, compared, say to chemotherapy for cancer. I don't think that such comparisons can be made. Instead, I'll give some details of a study of treatments for PTSD that is one of the better done studies: Foa et al., 1999. I'll admit, my selection of this study reflects my own biases, given that I work with Dr. Foa.

The study had three active treament conditions and a waitlist condition. The treatments were exposure therapy (we call it PE), stress inoculation training (SIT), and then the combination of PE/SIT. Subjects were all women assault victims who had chronic PTSD. Treatment consisted of 9 2-hour therapy sessions scheduled 2x per week. So the entire course of therapy was about 5 weeks long.

How did patients fare after treatment? All of the patients in the WL group still had PTSD. By contrast, 60% of subjects who received PE lost the diagnosis, as did 42% in SIT and 40% in the combined treatment. Now, these figures include not only people who completed treatment, but also people who dropped out of treatment early. Sometimes when people drop out, they don't come back for a second evaluation. When that happens, you substitute their pretreatment score for their posttreatment score. This is called an "intent-to-treat" analysis. For this particualr analysis, it is assumed that if you didn't return for an assessment, you still had PTSD, which may or may not have been the case. Moreover, if you stopped therapy very early on, it wouldn't be too surprising if you still had PTSD because you didn't get the full course of treatment. It would be like only taking half of your antibiotic to fight an infection.
What would these numbers look like if we limited ourselves to the people who actually completed treatment and have a real posttreatment score? This is called a "completers" analysis. In the Foa et al. study, 65% of people recieving PE lost the diagnosis of PTSD, as did 58% in SIT, and 54% in PE. Again, none of the waitlist subjects lost the diagnosis.

Now, losing the diagnosis of PTSD is all well and good, but in some cases it can literally boil down to the presence or absence of one symptom. This is due to the nature of how PTSD is diagnosed. A much higher bar is to assess "good endstate functioning." To achieve this, you need to show a low level of PTSD symptoms AND a score on a measure of general anxiety that is within normal levels AND a score on a measure of depression that is within normal levels. In other words, you have to feeling pretty darn good. For the intent-to-treat analysis, 52% of people receiving PE met this criteria, as did 31% in SIT, 27% in PE/SIT, but none in waitlist. Notice here that the combined treatment is not better than the individual treatments? More isn't neccessarily better. For the completers, the percent scores were 57, 42, 36, and 0 for PE, SIT, PE/SIT and waitlist, respectively.

If were were to look at studies of EMDR, cognitive therapy,and medication, we could find similar numbers, sometimes a bit higher, other times a bit lower.

All-in-all, most people get some benefit from treatment, about half to perhaps three quarters do pretty well, and somewhere between one third and one half do very well. By comparison, there is very little improvement in the absence of treatment. Our field can be proud of these achievements, but we also must be humbled by the fact that some people still don't do so well and we need to figure out why and how to help them. PTSD can be very chronic and cause extreme impairment that prevents people from functioning very well in their jobs, their roles as parents, their ability to enjoy recreational activities, and their ability to have close relationships. Good treatment can help to alleviate a good bit of this, but we still have a ways to go.

With that, I'll let you judge for yourself how well or how poorly our current treatments are doing.

Replies:

There are no replies to this message.

| Behavior OnLine Home Page | Disclaimer |

Copyright © 1996-2004 Behavior OnLine, Inc. All rights reserved.