Skip to main content
. 2005 Feb 19;330(7488):385. doi: 10.1136/bmj.330.7488.385

Table 1.

Summary of clinical trial data abstracted from the Medicine and Healthcare products Regulatory Agency's review of the safety of SSRIs9

SSRI (conditions included in RCTs; No of trials contributing data)
Active (SSRI) arm
Placebo arm
No of subjects No of episodes No of subjects No of episodes
(a) Suicides in placebo controlled trials in adults
Citalopram (depression; 9 trials) 1320 1 622 1
Escitalopram (all indications; 34 trials) 2648 1* 2088 1
Fluoxetine
Fluvoxamine (all indications; 48 trials) 4186 2 3396 2
Paroxetine (all indications; 95 trials) 8481 1 5808 3
Sertraline (all indications; 156 trials) 7169 4 5108 0
Total 23 804 9 17 022 7
Pooled odds ratio from bayesian random effects meta-analysis: 0.85 (95% credible interval 0.20 to 3.40)


(b) Non-fatal self harm in placebo controlled trials in adults
Citalopram (depression; 9 trials) 1320 11 622 5
Escitalopram (all indications; 34 trials) 2648 6 2088 1
Fluoxetine (all indications; 135 trials)§ 7010 17 4667 11
Fluvoxamine (all indications; 48 trials) 4186 24 3396 10
Paroxetine (all indications; 95 trials) 8481 33 5808 26
Sertraline (all indications; 156 trials) 7169 20 5108 8
Total 30 814 78 (+ 33 paroxetine) 21 689 35 (+26 paroxetine)
Pooled odds ratio from bayesian random effects meta-analysis: 1.57 (95% credible interval 0.99 to 2.55; 1.29, 0.90 to 1.91, with paroxetine data included)

(c) Suicidal thoughts in placebo controlled trials in adults
Citalopram (depression; 9 trials) 1320 10 622 4
Escitalopram (all indications; 34 trials) 2648 1 2088 2
Fluoxetine (all indications; 135 trials) 3078 24 1800 31
Fluvoxamine (all indications; 48 trials) 4186 23 3396 12
Paroxetine (all indications; (95 trials) 8481 32 5808 26
Sertraline (all indications; 156 trials) 7169 6 5108 6
Total 26 882 64 (+32 paroxetine) 18822 55 (+26 paroxetine)
Pooled odds ratio from bayesian random effects meta-analysis: 0.77 (credible interval 0.37 to 1.55; 0.79, 0.48 to 1.28, with paroxetine data included)
*

Suicide occurred six days after stopping treatment.

Data on suicide and self harm were not reported separately for fluoxetine.

All three placebo suicides occurred in the post-treatment period.

§

Data on suicide and self harm were not reported separately for fluoxetine, so a few events listed here may be suicides.

Data for non-fatal self harm and suicidal thoughts not reported separately for paroxetine, so events assumed to be evenly distributed across these two outcomes.