PSYCHOSOCIAL TREATMENT OF SCHIZOPHRENIA - COMMENTARY

THE preceding articles by Sandin and by Levander and Cullberg describe intensive and interesting work with seriously ill patients, but they also represent several elements of a general problem that stimulates this critique. Only one reference cited work other than that of the authors, but the subject matter has already received extensive attention in a rich and redundant literature. The task of integration of various therapeutic approaches was not addressed, although psychotherapy with the case described by Levander and Cullberg evidently took place in the context of respect for the other elements of treatment. Rules of evidence were not an apparent concern. Sandin does not inform the reader as to how or why cases were chosen, or how specific material was selected. Levander and Cullberg give quite a good description of their method for selecting a case. They sought successful cases of psychotherapy, but only 8 in 200 putative successes met their criteria. Rather than concluding that psychotherapy was not inducing the hoped for success, they selected one case with good outcome to illustrate elements of successful work. This was nicely done and even included a follow-up note of importance. But was the psychotherapeutic work causally related to outcome? The authors write in this explanatory framework that they sought reports of ''successful psychotherapeutic treatment.'' They interpreted data in addressing the question ''What were the characteristics of this seemingly successful therapeutic relationship?'' These authors are clear in purpose and provide new data by incorporating the patient's perspective. The language of presentation is that of explanation and cause, but the data, unsuited for that purpose, would be better applied to understanding and study of that which is meaningful in the therapeutic relationship. The only data relevant to psychotherapy as a cause of the observed outcome is the circumstantial evidence of looking at reported successful cases and finding so few that meet criteria. This, together with clinical trials testing psychodynamic psychotherapy efficacy (Fairweather et al. 1960; Grinspoon et al. 1972; Karon and O'Grady 1969; Karon and VandenBos 1970; May 1968; Rogers et al. 1967; Stanton et al. 1984), might have dissuaded the authors from seeking information on treatment success from this data base.