Cognitive—Behavior Therapy Versus Pharmacotherapy: Now That the Jury's Returned Its Verdict, It's Time to Present the Rest of the Evidence Neil S. Jacobson Department of Psychology University of Washington Steven D. Hollon Department of Psychology Vanderbilt University ABSTRACT Recent reanalyses suggest that pharmacotherapy was superior to cognitive—behavior therapy in the acute treatment of more severely depressed outpatients in the National Institute of Mental Health Treatment of Depression Collaborative Research Program (TDCRP). At the same time, this finding was neither robust across sites within the TDCRP nor consistent with findings from other studies. D. F. Klein has argued that those other studies were inherently flawed because they did not include pill-placebo controls, an argument that he extended to drug—psychotherapy comparisons in the treatment of panic as well. It is agreed that the inclusion of such controls would have facilitated the interpretation of the findings, but it is not agreed that their omission rendered those studies uninterpretable. Cognitive—behavior therapy remains a viable alternative to pharmacotherapy in the treatment of depression and a particularly promising intervention in the treatment of panic disorder. Preparation of this article was supported by a Research Career Development Award from the National Institute of Mental Health. Correspondence may be addressed to Neil S. Jacobson, Department of Psychology/CCR, University of Washington, 1107 Northeast 45th Street, Suite 310, Seattle, Washington, 98105. Received: June 29, 1995 Revised: September 1, 1995 Accepted: September 4, 1995 We have grown increasingly concerned about what we perceive to be a partisan debate among various interest groups who have become zealous in their advocacy of their preferred treatment models. Although science purports to serve the acquisition of knowledge and the pursuit of truth, it is all too easy to fall into the trap of interpreting data selectively in the service of an a priori position. This is especially likely to occur when the data from relevant studies are available only in summary form, because anything less than full disclosure of all the pertinent findings and the contexts surrounding their generation leaves undue latitude for preexistent beliefs to color the interpretive process. We think that proponents of both drugs and psychotherapy have gone beyond the data in drawing inferences about the efficacy (and effectiveness) of their respective interventions, thus transforming the pursuit of knowledge into a debate in which research findings are used selectively and interpreted inappropriately. At one extreme, there are advocates for pharmacotherapy who attempt to invalidate Journal of Consulting and Clinical Psychology © 1996 by the American Psychological Association February 1996 Vol. 64, No. 1, 74-80 For personal use only--not for distribution. Page 1 of 11Cognitive—behavioral treatment of panic disorder: Considerations for the treatment of patie12/13/2000http://spider.apa.org/ftdocs/ccp/1996/february/ccp64174.htmlevidence in support of psychosocial treatments and, in doing so, often distort or ignore findings that are inconsistent with their claims. At the other extreme, there are advocates for the psychotherapies who are guilty of a rush to judgment about the worth of the psychosocial interventions. We think that this has happened with respect to the treatment of depression and panic, two of the literatures that we know best, and we focus our discussion accordingly. At the same time, we suspect that we could have raised the same concerns with respect to the treatment of virtually any other disorder where drugs and psychotherapy have been compared. A Premature Postmortem for Cognitive—Behavior Therapy: The Treatment of Depression Collaborative Research Program In a recently published set of treatment guidelines prepared for the American Psychiatric Association, it was stated that cognitive—behavior therapy (CBT) may be no more effective than pill-placebo controls in the treatment of severe depression ( American Psychiatric Association, 1993 ). A similar conclusion was reached by the Agency for Health Care Policy and Research (AHCPR), which suggested that psychotherapy alone was inappropriate in the treatment of more severe depressions ( Depression Guideline Panel, 1993 ). This perspective has been widely publicized, in part because it corresponds so closely to preconceived notions held by many in the psychiatric treatment community. Unfortunately, it is based largely on results from a single study, which may not be as internally coherent as it first appeared: the Treatment of Depression Collaborative Research Program (TDCRP: Elkin, Parloff, Hadley, & Autry, 1985 ), sponsored by the National Institute of Mental Health (NIMH). The TDCRP was one of the largest and most ambitious treatment outcome studies ever conducted within the field of psychiatry. In that study, 250 unipolar depressed outpatients at three different sites were randomly assigned to one of four treatment conditions; CBT, imipramine plus clinical management (ICM), interpersonal psychotherapy (IPT), and a pill-placebo control. The initial report of the findings indicated few differences at posttreatment among the four conditions in the full sample ( Elkin et al., 1989 ). However, when the sample was split in terms of initial severity of depression and impairment, both ICM and IPT produced significantly greater acute response than did pill-placebo among the more severe subsample, whereas CBT did not. Advocates for pharmacotherapy have seized on the fact that drugs were superior to pill-placebo among the more severe patients but that CBT was not, whereas advocates for the psychotherapies (and particularly CBT) have argued that differences between drugs and CBT were neither significant nor robust across measures. The issue has been further complicated by an ongoing debate over whether tests for Treatment × Severity interactions should be considered a priori or post hoc and the degree of control that needs to be exercised over experimentwise error rates. Klein (1990) has argued that the TDCRP investigators chose too stringent a level of significance and that this conservatism led them to understate the superiority of pharmacotherapy over psychotherapy among the more severe subsample. Moreover, he claims to have alerted the TDCRP investigators to this potential
View Full Document
Unlocking...