L’évaluation du risque de violence est une démarche à la fois de plus en plus incontournable dans la psychiatrie médico-légale et à la fois fortement critiquée sur les plans éthique, sociétal et scientifique. Dans cet article, nous retraçons les différentes générations d’outils d’évaluation du risque de violence. L’évaluation structurée du risque a été créée en réponse aux manquements du jugement clinique non structuré. Il a été démontré dans maintes recherches antérieures que le jugement clinique non structuré comportait différents biais (surévaluation du risque, incohérence intercotateur, etc.). L’utilisation des échelles d’évaluation du risque répond en partie à ces manquements, mais comporte d’autres défauts (dangers d’utilisations erronées, utilité clinique incertaine, etc.). En analysant la littérature issue de ces différentes perspectives, nous décrivons les enjeux qui pèsent sur l’évaluation du risque de violence. Enfin, nous décrivons plusieurs pistes pour maintenir une attitude résolument clinique tout en intégrant ces outils d’évaluation dans la pratique.

Describe how (routine) structured risk assessment can be used in clinical practice without putting clinical commitment in jeopardy.

Violence risk assessment has become unavoidable in forensic psychiatry. Although it has become a routine task in many clinical settings, it has been criticized on ethical, societal and scientific grounds. In this article, we analysed literature stemming from different perspectives in order to present the both current and historical debates surrounding violence risk assessment. We remind that structured risk assessment was elaborated because of the many flaws of unstructured clinical judgement (overestimation of risk, lack of interrater agreement, etc.). On the other hand, structured risk assessment has its own flaws (risk of misuses, uncertain clinical usefulness, etc.). We feel that this scientific debate has also been complexified by the many fears and resistances of clinicians, especially in the Roman-European traditions.

Literature review from different literature threads. We review historical, empirical and statistical, clinical and philosophical literature on the subject of risk assessment.

We argue that every new technology (in this case, structured risk assessment) brings about its own share of concern and suspicion, but that careful appraisal is better than steep resistance. To this end, we describe methods and perspectives on how to use structured risk assessment without compromising clinical priorities. For example, specifying the nature of the assessed risk to decrease confusion and amalgamation, communicating about risk assessment inferences with patients, using risk assessment data as a third in clinical discussions, developing the capacity to translate empirical research findings into clinical practice, clarify the advantages of routine monitoring, etc.