Weighing up risks and benefits of antidepressants
A multiple-treatments meta-analysis comparing 12 new-generation antidepressants has recommended sertraline as the drug of choice on the grounds of efficacy, acceptability and cost.
A total of 117 randomised controlled trials (25, 928 participants) met the inclusion criteria for the meta-analysis.
The response rate was chosen as a measure of efficacy, defined as the proportion of patients with a depression rating scale score at least 50% below the baseline score after 8 weeks of treatment.
The dropout rate was chosen as a measure of acceptability, defined as the proportion of patients who discontinued their treatment within 8 weeks.
Mirtazapine, escitalopram, venlafaxine and sertraline were ranked as the four most effective antidepressants, with significantly greater efficacy than duloxetine, fluoxetine, fluvoxamine, paroxetine and reboxetine. The four best tolerated antidepressants were escitalopram, sertraline, bupropion and citalopram. Escitalopram and sertraline combined both efficacy and acceptability, and sertraline was chosen as a ‘best buy' on account of its lower cost.
Multiple-treatments meta-analysis incorporates both direct (within a trial) and indirect comparisons (between trials which have used the same comparator treatment). This may help to reduce, but will not eliminate, sponsorship bias. Only 15 of the studies included were unpublished (13%) as compared with 31% of FDA-registered antidepressant trials,1 which suggests that the results are likely to have been influenced by selective publication bias.
The use of an arbitrary cut-off value to distinguish responders from non-responders can magnify the apparent difference between two treatments: a large difference in the response rates can result from a small difference in the average percentage change in score.2 One consequence of this may be to magnify the bias in favour of antidepressants with greater sedative properties, which results from the fact that, out of a maximum score of 52 on the Hamilton depression rating scale, there are 12 points for agitation/anxiety and 6 points for insomnia.3
There are factors other than short-term efficacy and acceptability that we need to bear in mind when selecting an antidepressant. There are important differences in the risk of suicide: a large cohort study from Finland found that fluoxetine had the lowest risk (relative risk 0.52) whereas venlafaxine had the highest risk (RR 1.61).4 Antidepressant discontinuation syndrome is more likely with drugs with relatively short half-lives, such as paroxetine, than in those with longer half-lives, such as fluoxetine.5
My current first choice antidepressant is citalopram (moderate efficacy, high acceptability, low suicide risk, intermediate half-life). While I might be tempted to consider switching to sertraline, I am not yet convinced that this would achieve a clinically significant benefit.
Cipriani A, Furukawa TA, Salanti G et al. Comparative efficacy and acceptability of 12 new-generation antidepressants: a multiple-treatments meta-analysis. Lancet 2009;373:746-58Reviewer
Dr Phillip Bland