I was confused by your October/November issue. In Nikolas Westerhoff’s article “Fantasy Therapy” I read that psychologists “treated male disaster workers traumatized by the World Trade Center attacks of September 11 by exposing them to realistic renditions of planes flying over virtual twin towers....” But then in “Brain Stains” I read, “For example, research ... has shown that reliving traumatic memories shortly after a terrifying event—performed in a popular therapeutic technique called crisis debriefing—may cause unnecessary stress and impede recovery.”
Are some traumas so damaging that once they have occurred there is not much therapy can do?
LAMBERT AND LILIENFELD REPLY: Regarding the question of when, if ever, therapeutic exposure to traumatic experiences is helpful, both learning theory and scientific evidence offer guidance. Exposure can be helpful, but only when it is sufficiently prolonged to permit clients’ anxiety to dissipate. One of the key shortcomings of crisis debriefing is that it is typically conducted in an uncontrolled fashion—some clients may leave sessions less anxious than when they entered, whereas others may leave sessions more anxious. For the latter individuals, crisis debriefing may be harmful.
MOOD MEDS VS. PLACEBOS
“The Best Medicine?” [Facts and Fictions in Mental Health], by Hal Arkowitz and Scott O. Lilienfeld, is a valuable article on the advantages of cognitive-behavior therapy over antidepressants. But the authors err in repeating the highly inflated claim of 67 percent effectiveness for antidepressants in the study by psychiatrist A. John Rush and his colleagues, which offered patients a four-step sequence of different antidepressant medications. If patients did not attain remission at one stage, they could then try a different antidepressant.
It is important to note that this study included no placebo control groups. Published studies that do include such controls typically find a 25 to 30 percent success rate with placebo. Only one drug in Rush’s study achieved even that rate of remission—all other drugs and drug combinations did worse. Rush’s 67 percent figure came from cumulating across trials without taking into account the placebo effects operating within each trial.
Moreover, supporting the explanation that antidepressants provide primarily a placebo effect, patients showed very high relapse rates consistent with the time-limited value of placebos.
The widely quoted 67 percent figure is bogus. I am the second author of an article soon to be submitted for publication that provides a critique of this study, citing the placebo problem as well as other issues.
Allan M. Leventhal
Silver Spring, Md.
ARKOWITZ AND LILIENFELD REPLY: Leventhal raises two different but related questions about Rush’s study. The first is, How effective was the treatment sequence used in this study (regardless of what was responsible for its effectiveness)? We disagree that the 67 percent finding is “bogus.” Irrespective of what caused this outcome (active medication or placebo effect), it is true that 67 percent of the patients were in remission by the end of the study.
The second question relates to the degree to which the outcomes could be attributed purely to the active effects of the medications. Leventhal correctly points out that the absence of placebo control groups in the study makes it unclear whether the outcomes were the result of the drugs or placebo factors such as expectations of change and supportive contact with the research staff. This issue becomes even more significant in light of the fact that most studies comparing antidepressants with placebos usually show only a small advantage for the medications.