Selective serotonin reuptake inhibitors (SSRIs) and suicide in adults: meta-analysis of drug company data from placebo controlled, randomised controlled trials submitted to the MHRA's safety review

The authors of this article published last year, David Gunnell and colleagues, have alerted us to an error in the abridged version of their paper (BMJ 2005;330:385-8). In the table, the correct estimate for the pooled odds ratio for self harm from the bayesian random effects meta-analysis for non-fatal self harm in relation to use of selective serotonin reuptake inhibitors (excluding paroxetine) is “1.57 (credible interval 0.99 to 2.55)”—not 1.51 (0.95 to 2.49). This matches the values given in the abstract and in the results section of the paper.