I hope so. Is there a science to how to make the most of our interactions with our patients? Probably. Does our enthusiasm sometimes outstrip our data? More frequently than we would care to admit—we are human, after all.
Cognitive behavior therapy for depression has been, perhaps, the most successful psychosocial treatment ever introduced. It has been the subject of dozens of clinical trials and multiple collaborative studies and is the intervention most likely to be taught in psychiatric training programs. Why, then, if the authors of this month’s article on cognitive behavior therapy for schizophrenia (Turkington et al.) are correct, has there been such a low level of interest in the United States in making cognitive behavior therapy a first-line intervention in schizophrenia?
For one of the few times in the history of individual psychosocial treatments of schizophrenia, we have the possibility of a treatment that may well interact with medication, possibly extending the effectiveness of it and certainly not seen as a direct competitor to it. The treatment itself offers an alternative, compassionate approach to people with schizophrenia without having to accept a disease-based formulation. It is reminiscent of the Kingsley Hall/Soteria approach (1, 2) without the antimedication element. Cognitive behavior therapy attempts to integrate the concept that psychosis is derived from a series of stress-induced misinterpretations into the patient’s way of thinking. It appears to be possible to do this without endorsing a diagnostically driven, medical model. Should this approach supplant the medical model and the discussion of a diagnosis? I will leave that to a much longer editorial, which I hope someone else will be asked to write.
We can all agree that cognitive behavior therapy is a reasonable alternative treatment. The authors call for further, urgently needed research. I certainly agree. Their characterization of the United States as having a medication-driven treatment agenda is also easy to understand. The results of the Clinical Antipsychotic Trials of Intervention Effectiveness have just been published (3). Although we will probably never have a full accounting of the resources needed to complete that study (most of the patients could not complete the study), it would not be much of a stretch to conclude that resources spent in one treatment direction cannot be spent in another. Psychosocial treatment research has fed on "table scraps" for too long. We can be deeply grateful to Turkington et al. for pointing out that the relative neglect of psychosocial treatment research is not for want of promising, new ideas and data that, if not sufficient to convince, are at least sufficient to drive our scientific need to go further.
The authors hypothesize that cognitive behavior therapy for schizophrenia has not been embraced in the United States because of the development of a "PTSD-like" state among U.S. mental health practitioners for any individual psychotherapy for schizophrenia. They posit that the 30-year war between psychoanalysis and biological psychiatry—won, in their minds, by the biologists—has left clinicians in the United States reluctant to apply individual psychotherapy to schizophrenia. Their stance would appear to be critical of the United States for not being as accepting of cognitive behavior therapy for schizophrenia as our colleagues in the United Kingdom have been. I would like to offer an alternate explanation of what may be driving a slow uptake of acceptance of cognitive behavior therapy in the United States.
Treatment acceptance is driven by clear, consistent, replicable data. The data cited by Turkington et al. do not completely establish clear, consistent, replicated outcomes—there is a great deal to suggest that cognitive behavior therapy could be a successful treatment but less to prove it.
The two flagship studies exemplify the problems with consistency. Tarrier et al. (4–6) found cognitive behavior therapy to have an advantage over supportive counseling in the short term (10 weeks) but not in the long term (1–2 years). Sensky et al. (7), on the other hand, found no short-term (9 months) advantage to cognitive behavior therapy compared with "befriending" therapy but found a long-term advantage (subsequent 9 months). Similar inconsistencies occur when medication adherence is used as an outcome. Kemp et al. (8, 9) found an advantage for cognitive behavior therapy; O’Donnell et al. (10) did not. Turkington et al. cite the work of McGorry (11) as providing evidence that cognitive behavior therapy may have a role in preventing or delaying transition from a preschizophrenia state to a full-blown diagnosis of schizophrenia. Indeed, when McGorry compared a specific intervention (low-dose atypical antipsychotic plus cognitive behavior therapy) to a needs-based psychosocial intervention, the specific intervention was significantly better at 6 months. However, when he then divided the patients in the specific intervention into those who were fully or not fully compliant with their medication, a somewhat different result occurred. Patients receiving cognitive behavior therapy who were not compliant with medication did no better than those receiving needs-based intervention alone—the medication appears to account for most of the outcome variance.
The conclusions of Turkington et al. on the effectiveness trials is that they "generally favor the cognitive behavioral intervention, albeit less strongly than some of the predecessor cognitive trials." The authors also propose a number of what appear to me to be straw men in support of cognitive behavior therapy and the system that endorses it in the United Kingdom: 1) Poorly conducted psychoeducation is dangerous—most poorly conducted endeavors are. 2) The U.S. preference for a treatment approach that is linked to a diagnosis—without a diagnosis, it seems much harder to build a case for continuing antipsychotic medication during symptom-free intervals, and the research literature is clear that medication discontinuation leads to relapse. 3) The United States has a singular reliance on antipsychotic medication—do we not even talk to our patients in the United States, and yet, who among us would treat schizophrenia without medication?
The authors conclude that "the literature on cognitive behavior therapy has been favorable enough to make use of this intervention for schizophrenia a treatment recommendation in the United Kingdom. However, the evidence from the literature is not definitive." Relying on enthusiasm, evangelical beliefs, anecdote (the plural of which is not data), or even questionable data does not sway opinions in those who accept scientific methodology as a cornerstone of medicine. In mental health, we have generations of claims that did not work, patients who have suffered because of them, and precious research resources wasted. It is not just psychosocial treatments such as psychoanalysis for schizophrenia that have been the repository of beliefs trumping science. One need only to go back to the 1970s to find hemodialysis (12) and beta-endorphins (13) as treatments for schizophrenia gracing the pages of our most prestigious journals.
To turn a phrase from another English author whose ideas have found considerable acceptance in the United States: I come not to bury cognitive behavior therapy, but to praise it. Cognitive behavior therapy is a promising treatment, but in our need to find new, better treatments for schizophrenia, let us avoid overpromising. Still to be more fully developed are issues of which patients, at which stage of illness, for how long, with what essential components, seen by therapists with what training, and at what costs demand to be answered. Our patients can look forward to the future explication of cognitive behavior therapy in schizophrenia.
Address correspondence and reprint requests to Dr. Keith, Department of Psychiatry, University of New Mexico School of Medicine, 2400 Tucker NE, Rm. 404, Albuquerque, NM 87131; SKeith@salud.unm.edu (e-mail).