Published online by Cambridge University Press: 20 July 2007
The association between treatment-emergent suicidality as an adverse event and fluoxetine treatment was examined using a fluoxetine double-blind placebo-controlled database of clinical trials for indications other than major depressive disorder.
The database consisted of 53 trials for 16 different indications (14 psychiatric, two non-psychiatric). Within each study, patient adverse event reports and narratives were searched extensively for treatment-emergent thoughts and behaviors associated with suicide. The incidence of adverse events was classified using Food and Drug Administration (FDA) codes for completed suicide, preparatory acts, suicidal ideation and the summary category of ‘all suicidality.’ The risk difference and risk ratios between fluoxetine and placebo treatment arms were compared using Mantel–Haenszel methods.
Within this large database, patients were randomly assigned to receive treatment with either fluoxetine (n=7066) or placebo (n=4382). Treatment groups did not differ in their risk for the emergence of suicidality for any FDA code; the risk ratio for ‘all suicidality’ was 0·82 (p=0·406), and there were no completed suicides in either group. Analyses based on treatment indication (bulimia, obsessive-compulsive disorder, other psychiatric and non-psychiatric illness) also showed no significant difference in risk between treatment groups. When examined by age categories (18–24, 25–30, 31–65, and ⩾65 years), fluoxetine and placebo treatments did not result in significant risk difference for the emergence of suicidality.
The risk of treatment-emergent suicidality does not appear to be associated with fluoxetine treatment for adults with various non-MDD conditions.