Special feature: personality disorders: models for treatment and strategies for treatment development.

A critical review of the literature suggests that the only approaches specific to the treatment of personality disorder that have been proven to be effective are behavioral, cognitive behavioral, or psychopharmacological. Those investigations have used a conventional protocol for clinical outcome studies commonly used in medical science. This protocol is described in a listing of Empirically Validated Therapies (EVT). The design calls for relatively brief treatments that are clearly operationalized, and studied in terms of symptom change in a treatment group relative to an appropriate clinical contrast group. It is argued here that such relatively brief studies of symptom or specific behavioral change do not address what is meant by personality disorder. The prevailing EVT model is challenged in two ways: (1) Efficacy of the approaches that have followed this model, namely the behavioral and psychopharmacological treatments, may have been overstated. (2) Because they are so brief, and because they assess symptoms, rather than relevant dynamic concepts like unconscious wishes and fears that organize personality, these protocols seem inappropriate to the study of effective treatment of personality disorder per se. Models that would be appropriate to treatment of personality disorder require testable causal theory rationally linked to treatment interventions, therapy process, and outcome measures. Some relevant modifications to research protocol for psychopharmacological, as well as psychosocial treatments, are suggested. A therapy approach that could meet these standards for testable causal theory is sketched.