Cette étude montre que les corrélations entre les échelles d’idéation suicidaire de Beck et d’évaluation du risque suicidaire de Ducher (RSD) sont statistiquement significatives (r=0,69 ; p<0,0001) mais que ces deux échelles ne sont pas mieux corrélées avec l’échelle Hamilton dépression qu’avec l’échelle d’anxiété clinique de Snaith. De plus, l’utilisation d’un modèle multifactoriel avec sélection en step-wise révèle que l’item « suicide » de l’échelle Hamilton dépression explique, à lui seul, plus du tiers de la variabilité de l’échelle de Beck et de la RSD. Ainsi, la relation entre les dimensions « suicide » et « anxiété » apparaît aussi importante chez notre population de patients dépressifs que celle entre les dimensions « suicide » et « dépression ». L’amélioration plus rapide, sous fluoxétine ou sous fluvoxamine, de la RSD face à l’échelle de Beck ou de Hamilton et l’évolution parallèle de ces dernières suggèrent que certains antidépresseurs pourraient avoir une rapidité d’action sur la dimension « suicide » comparable à celle qu’ils ont sur la dimension « dépression ». Le fait que l’échelle d’évaluation du risque suicidaire objective une différence statistiquement significative entre les deux groupes de traitement renforce l’hypothèse d’une inégalité des antidépresseurs face au risque suicidaire.

Most of the people who will attempt suicide, talk about it beforehand. Therefore, recognition of suicidal risk is not absolutely impossible. Beckʼs suicidal ideation scale and Ducherʼs suicidal risk assessment scale (RSD) are common tools to help practicians in this way.

These scales and the Hamiltonʼs depression scale were included in an international multicentric, phase IV, double-blind study, according to two parallel groups who had been administered a fixed dose of fluvoxamin or fluoxetin for six weeks. This allowed examination of the correlations between these scales and the relations, which could possibly exist between suicidal risk, depression and anxiety.

(a) Relationships between the Beckʼs suicidal ideation scale, the suicidal risk assessment scale RSD and Hamiltonʼs depression before treatment. Before treatment, the analysis was conducted with 108 male and female depressive outpatients, aged 18 or over. Results revealed a significant positive correlation (with a Pearsonʼs correlation coefficient r equal to 0.69 and risk p<0.0001) between Beckʼs suicidal ideation scale and the suicidal risk assessment scale RSD. These scales correlate less consistently with Hamiltonʼs depression (Beck/Hamiltonʼs depression: r=0.34; p=0.0004—RSD/Hamiltonʼs depression: r=0.35; p=0.0002). We observed that the clinical anxiety scale by Snaith is also strongly correlated to these two suicidal risk assessment scales (Beck/CAS: r=0.48; p<0.0001—RSD/CAS: r=0.35; p=0.0005). Besides, the item “suicide” of Hamiltonʼs depression scale accounts for more than a third of the variability of Beckʼs suicidal ideation scale and the suicidal risk assessment scale RSD. According to these results, the suicidal risk evaluated by these two scales seems to be significantly correlated with anxiety as much as with depression. On the other hand, the Clinical Global Impression is fairly significantly correlated with Beckʼs suicidal ideation scale (r=0.22; p=0.02), unlike the suicidal risk assessment scale RSD (r=0.42; p<0.0001) and Hamiltonʼs depression scale (r=0.58; p<0.0001); (b) Relationships between Beckʼs suicidal ideation scale, the suicidal risk assessment scale RSD and Hamiltonʼs depression under treatment. The follow-up under treatment (fluvoxamin or fluoxetin) during six weeks revealed the significantly better sensitivity of the RSD in comparison with Beckʼs suicidal ideation scale and Hamiltonʼs depression scale, showing the significantly faster improvement in the RSD (p<0.0001). There was no significant difference between the evolution of Beckʼs suicidal ideation scale and Hamiltonʼs depression scale. So, under treatment with fluvoxamin or fluoxetin, the improvement in suicidal risk appears to be as rapid as the improvement in depression. If we look at the treatment prescribed, only the suicidal risk assessment scale RSD revealed a significant difference between the two molecules, with more rapid improvement with fluvoxamin (p=0.015) from D14.

In conclusion, the results of this study hypothesize that the suicidal risk, as assessed by Beckʼs suicidal ideation scale and the suicidal risk assessment scale RSD, appears to be consistently correlated with both the level of anxiety and depression. The study also suggests that all antidepressants may not be equally effective on suicidal risk.