Because psychoanalysis is a complex clinical discipline, practitioners do not have the opportunity to determine which of the many specific interventions used in work with individual patients are mutative agents responsible for observed therapeutic change. This state of affairs is illustrated in a dramatic form when we attempt to determine the interplay between psychodynamic and psychopharma-cological interventions in terms of therapeutic efficacy.

Despite this, in their work with highly symptomatic patients, all clinicians must arrive at a formulation for the optimal point at which a recommendation for medication should be made. No definitive guidelines exist for making this determination, and the decision involves consideration of multiple factors that include the patient's suicidality, level of symptomatology and dysfunction, and degree of co-morbid depressive symptoms, as well as the nature of the transferential meaning of the act of prescribing medication for the patient. This