EMDR for Panic Disorder With Agoraphobia Comparison With Waiting List and Credible Attention-Placebo Control Conditions Alan J. Goldstein Department of Psychology University of North Carolina at Chapel Hill Edwin de Beurs Department of Psychology University of North Carolina at Chapel Hill Dianne L. Chambless Department of Psychology University of North Carolina at Chapel Hill Kimberly A. Wilson Department of Psychology University of North Carolina at Chapel Hill ABSTRACT In a randomized controlled trial, eye movement desensitization and reprocessing (EMDR) for panic disorder with agoraphobia (PDA) was compared with both waiting list and credible attention-placebo control groups. EMDR was significantly better than waiting list for some outcome measures (questionnaire, diary, and interview measures of severity of anxiety, panic disorder, and agoraphobia) but not for others (panic attack frequency and anxious cognitions). However, low power and, for panic frequency, floor effects may account for these negative results. Differences between EMDR and the attention-placebo control condition were not statistically significant on any measure, and, in this case, the effect sizes were generally small (η 2 = .00—.06), suggesting the poor results for EMDR were not due to lack of power. Because there are established effective treatments such as cognitive—behavior therapy for PDA, these data, unless contradicted by future research, indicate EMDR should not be the first-line treatment for this disorder. Although a relatively new and still controversial treatment, eye movement desensitization and reprocessing (EMDR; Shapiro, 1995 ) has spawned research concerning its efficacy and active elements, the bulk of this on posttraumatic stress disorder (PTSD). Much of this research is methodologically flawed (see Feske, 1998 ); however, there are a few well-designed studies documenting the superiority for PTSD of EMDR to a waiting list control group ( Rothbaum, 1997 ; Wilson, Becker, & Tinker, 1995 ) or, in one case, to supportive listening ( Scheck, Schaeffer, & Gillette, 1998 ). After a rigorous review of this literature, several authors ( Chambless et al., 1998 ; DeRubeis & Crits-Christoph, 1998 ; Feske, 1998 ) concluded that EMDR appears to be beneficial for civilian PTSD but that its efficacy for combat-related PTSD remains to be substantiated in well-designed research. In a meta-analysis of PTSD studies, van Etten and Taylor (1998) found that, on most posttest measures of PTSD, EMDR was comparable to behavior therapy (a grouping of two effective treatments: exposure or stress inoculation training) and selective serotonin reuptake inhibitors (SSRIs) in efficacy and was superior to control conditions. Follow-up effect sizes for EMDR and behavior therapy continued to be equivalent. These results must be taken with caution because in no study included was EMDR directly compared with behavior therapy or SSRIs. Nonetheless, it is reasonable to conclude that EMDR offers some benefit for some forms of PTSD, at least when the trauma is not related to combat, even though the mechanisms for this efficacy remain in dispute ( Steketee & Goldstein, 1994 ). Journal of Consulting and Clinical Psychology © 2000 by the American Psychological Association December 2000 Vol. 68, No. 6, 947-956 For personal use only--not for distribution. Page 1 of 189/5/2001http://spider.apa.org/ftdocs/ccp/2000/december/ccp686947.htmlShapiro (1995) , who developed EMDR, has advocated its use for a wide variety of disorders other than PTSD, including other anxiety disorders. Uncontrolled case studies suggest potential benefits for panic and phobias (see, e.g., Goldstein & Feske, 1994 ; Marquis, 1991 ), but controlled research is sparse. One research group in the Netherlands is responsible for the bulk of EMDR for phobia research conducted with randomized controlled designs and sound measurement. 1 In two studies, Muris and Merckelbach and their colleagues compared EMDR with a waiting list ( Muris & Merckelbach, 1997 ) or attention-placebo control group ( Muris, Merckelbach, Holdrinet, & Sijsenaar, 1998 ) for treatment of spider phobia in adults or children, respectively. Treatments were limited to one 1- or 2.5-hr session. On behavioral avoidance tests, EMDR was not superior to the control conditions, but on a self-report measure of phobia (used only in the study with children), EMDR participants were significantly more improved than the control group children. Comparing EMDR with an assessment-only group for public-speaking anxiety (participants were not required to meet criteria for social phobia), Foley and Spates (1995) found one to two sessions of EMDR to be significantly more effective for two self-report outcome measures but not for heart rate or observable anxiety during a speech. However, the sample size was quite small (8—10 per group), and, for all measures except heart rate, the pattern of the data indicated more change in the EMDR than in the assessment-only group. Testing EMDR against an active treatment, Muris, Merckelbach, and colleagues ( Muris & Merckelbach, 1997 ; Muris et al., 1998 ; Muris, Merckelbach, van Haaften, & Mayer, 1997 ) compared one session of EMDR with one session of imaginal or in vivo exposure for spider phobia. Sessions were 1, 1.5, or 2.5 hr in duration, depending on the study. In general, in vivo exposure proved superior to EMDR on self-report measures and the behavioral avoidance test, with results sometimes significantly different and other times showing trends favoring exposure. The effects of one session of EMDR versus one session of imaginal flooding were not significantly different. Overall, these studies provide little support for EMDR's efficacy as a treatment of animal phobia. As the authors acknowledged, there are a number of methodological limitations to this research, including the questionable external validity of 1-hr treatment, failure to assess treatment integrity, and, in two studies, confounding of therapists and therapists' experience with treatment condition. Moreover, samples sizes were quite small (8 or 9 participants per group), yielding adequate power to detect only very large between-groups effects (e.g., Cohen's d of >=1.07). In a somewhat larger, more clinically representative study, Feske and Goldstein (1997) compared six sessions of EMDR ( n = 15) with a waiting list control condition ( n = 12) for panic disorder (almost always with agoraphobia).
View Full Document
Unlocking...