Quatre-vingts pour cent des sujets qui font un passage à l’acte expriment des idées suicidaires dans les mois qui précèdent. L’extériorisation de telles idées n’est donc pas un facteur protecteur, mais bien un facteur de risque à rechercher systématiquement. Plusieurs instruments d’évaluation cherchent à aider les cliniciens ou les chercheurs dans cette démarche. On peut en particulier citer l’échelle d’idéation suicidaire de Beck et l’échelle d’évaluation du risque suicidaire RSD. Leur inclusion dans une étude multicentrique internationale, en double aveugle, avec dose fixe pendant 6 semaines de fluoxétine ou de fluvoxamine, a permis d’étudier les corrélations existant entre elles. L’analyse avant mise sous traitement retrouve une corrélation satisfaisante entre la RSD et l’échelle de Beck (r = 0,69 ; p ≪ 0,0001), ainsi qu’entre la RSD et l’item « suicide » de l’échelle de dépression de Hamilton (r = 0,60 ; p ≪ 0,0001). En revanche, une corrélation moins satisfaisante existe entre la RSD et le score global de l’échelle de Hamilton (r = 0,35 ; p = 0,0002). Au cours de l’évolution sous traitement, l’amélioration de la RSD se montre significativement plus rapide que celle des 2 autres échelles (p ≪ 0,0001) et est influencée par le type de traitement (p = 0,015), ce qui soulève certaines questions. De plus, un score de 7 à la RSD semble représenter aux yeux des expérimentateurs un niveau de risque jugé particulièrement significatif.

The prevention of suicide is a top priority in mental health. The determination of high risk suicidal groups is not sufficient. The expressing suicidal ideas is not a protective factor, but in contrary a risk factor to take into account, or even to search and to quantify : 80 % of the subjects who attempt to commit suicide or commit suicide express such ideas months before. Several evaluation instruments try to help the practicians or the research workers in this reasoning. The suicidal risk assessment scale RSD can be cited in particular. It is composed of eleven sections. The 0 level corresponds to the absence of particular ideas of death or suicide. Levels 1 and 2, the presence of ideas of death. Levels 3-4-5, the presence of suicidal ideas. The difference compared to the majority of the other scales consecrated to the same subject, the passif desire of death, occupies a place totally particular in the RSD (level 6). From the level 7, the risk of acting out seems to become more important. It stops being a simple idea of suicide, but becomes a real will of dying, firstly retained by something or someone (level 7), the fear of causing suffering to dear ones or a religious belief…, then determinated (level 8). Finally, the patient has elaborated a concrete plan (level 9) or he has already started the preparation of acting out (level 10). It is just necessary to evaluate and to note the hightest level of the scale. The inclusion of the suicidal risk assessment scale RSD and of the Suicidal Ideation Scale by Beck in an international multicentrist, phase IV, double-blind study, according to two parallel groups, with a fixed dose of fluoxétine or fluvoxamine for six weeks, allowed to search correlations which could exist between the two scales. The analysis before the beginning of the treatment was done on 108 outpatients depressive, male and female, aged 18 or over. It finds a satisfactory concurrent validity between the suicidal risk assessment scale RSD and the Suicidal Ideation Scale by Beck (r = 0.69 ; p ≪ 0.0001) as well as between the RSD and the item « suicide » of the Depression scale by Hamilton (r = 0.60 ; p ≪ 0.0001). On the other hand, it is less satisfactory between the suicidal risk assessment scale RSD and the Hamilton Depression scale overall score (r = 0.35 ; p = 0.0002). During the evolution under the treatment, the suicidal risk assessment scale RSD shows an improvement significantly faster than the Hamilton Depression scale or the Beck’s Suicidal Ideation Scale (p ≪ 0.0001). This statement of fact arouses some questions about the suicidal risk of which the evolution in the case of a treated depressive episode could be quicker than first thought. All the more so as this difference is affected by the type of the treatment (p = 0.015). Moreover, a score of 7 and more on the suicidal risk assessment scale RSD seems to represent a risk level judged particularly significant by the experimenters. In effect, the existence of such a suicidal risk was a criteria of exclusion and no patient with a level superior to 6 on suicidal risk assessment scale RSD was included. In conclusion, the utilisation of the suicidal risk assessment scale RSD could be interesting in the prevention of suicide.