Jump to: Page Content, Site Navigation, Site Search,
You are seeing this message because your web browser does not support basic web standards. Find out more about why this message is appearing and what you can do to make your experience on this site better.
BMJ 2005;330:1149 (14 May), doi:10.1136/bmj.330.7500.1149-a
EDITORThe news media's preoccupation with whether antidepressants provoke suicidal behaviour has generated apprehension in the general public, and clinicians increasingly see patients resistant to taking selective serotonin reuptake inhibitors (SSRIs) even though they might benefit. The meta-analysis by Fergusson et al reported an excess of suicidal attempts with SSRIs v placebo and, initially, an alarming sevenfold odds ratio for fatalities compared with tricyclic antidepressants [subsequently corrected to 1.08 (0.28 to 4.09), see correction 19 March, p 653].1
A statistical excess of suicidal attempts in studies of SSRIs v placebo has been previously reported and was reasonably explained in the accompanying editorial by Cipriani et al.2 3 However, the whole debate loses sight of the fact that the underlying trials were never designed to assess suicidality as an outcome but to satisfy regulatory agencies about efficacy. Retrospective counts of incidents of deliberate self harm or attempted suicide are extremely unreliable in such studies; fatalities obviously less so (but no excess has surfaced among these). The randomisation process in smaller trials may be questioned, and heterogeneity could have precluded some of the trials in the current meta-analysis. Prospective studies with suicidality as the outcome variable are needed to lay such issues to rest, but these have rarely been done, for valid reasons.4
In any case, odds ratios (or other ratios) alone do not give an indication of absolute risk: the number needed to harm (NNH) should also be examined. In the current meta-analysis, an NNH of 708 compares quite favourably with others in medicinefor example, 179 in the CAPRIE trial (comparing clopidogrel with aspirin for stroke patients).5
Isaac Sakinofsky, professor emeritus of psychiatry and public health sciences
University of Toronto, CAMH-Clarke Site, 250 College Street, Toronto, ON, Canada M5T 1R8 isaac.sakinofsky{at}utoronto.ca
David L Streiner, director
Kunin-Lunenfeld Applied Research Unit, Baycrest Centre for Geriatric Care, Toronto, ON, Canada M6A 2E1
Israeli students are refusing to perform intimate examinations on anaesthetised women without their informed consent.