[Contributions of epidemiology to therapeutic decisions in psychiatry].

Since the 70's, the study of interventions methods (evaluation of their influence upon disease and of their use by care-giving structures) has been one of the main fields of research in epidemiology. Consequences of these studies upon therapeutic decisions can be the following: 1) From the diagnostic point of view, epidemiology has much contributed to the elaboration of the most recent classification systems (DSM III and IV, ICD 10) and to the determination of detection strategies in order to improve knowledge about psychiatric disorders in primary care. 2)The contribution of epidemiology to the decision to treat varies according to the kind of intervention: epidemiology has brought definitive means in evaluating prognosis of disease, the basis of tertiary care (prevention of relapse, recurrence and chronicity); epidemiology works have also identified many risk factors for various mental diseases; they sometimes allow to define secondary prevention strategies about a precise risk factor, but nevertheless have not led to elaborate predictive models in order to identify subjects candidate to specific treatments. 3) Epidemiology has also contributed to improve choice among treatments, depending on an advantage/risk ratio: as the first advantage expected from a treatment, efficacy remains difficult to evaluate for some therapeutic decisions (hospitalization, psychotherapy,...). Experimental epidemiology though led to establish references in the pharmacological treatment field for many psychiatric disorders. Nevertheless, utilisation of results brings many problems, theoretical (interpretation and generalization of results), as well as practical (applicability); evaluation of risks due to medications (side-effects and accidents, suicide risk) has been sustained by pharmacovigilance studies, the results of which led in some cases to establish regulations for prescription.