A total of 315 trials were included in the review. Within these trials there were 355 paired comparisons(n=36,509). Most comparisons (206) were of newer with older antidepressants. There were 114 comparisons of newer antidepressants with placebo and 37 comparisons between newer antidepressants. There were 14 trials that compared St. John's Wort (hypericum) with either placebo (8 studies) or older antidepressants (6 studies), but none that compared hypericum with newer antidepressants.
The following 24 questions were asked in this review.
Are newer antidepressants more effective than placebo or older antidepressants for treating adult patients with major depression?
Are newer antidepressants more effective than placebo or older antidepressants for treating adult patients with dysthymia?
Are newer antidepressants more effective than placebo or older antidepressants for treating adult patients with mixed-anxiety depression?
Are newer antidepressants more effective than placebo or older antidepressants for treating adult patients with subsyndromal depressive disorders?
Are newer antidepressants more effective than placebo or older antidepressants for treating adult patients with recurrent depression?
Are newer antidepressants more effective than older antidepressants for treating adult patients with refractory depression?
Is hypericum more effective than placebo or standard antidepressants for treating depressive disorders in adults?
Are valeriana and kava more effective than placebo or standard antidepressants for treating depressive disorders in adults?
Are newer antidepressants more effective than placebo or older antidepressants for treating depressive disorders in children and adolescents?
Are newer antidepressants more effective than placebo or older antidepressants for treating older persons with depressive disorders? Are newer antidepressants more effective than placebo or older antidepressants for treating patients with co-morbidity?
Does the efficacy of newer agents vary between men and women and between different ethnic groups?
Are newer antidepressants more effective than placebo or older antidepressants for treating adult primary care patients with depressive disorders?
Are newer antidepressants more effective than placebo or older antidepressants in the postpartum setting?
Are combinations of newer antidepressants with other antidepressants more efficacious than a single antidepressant for treating major depressive disorder in adults?
Are combinations of newer antidepressants with other antidepressants or anxiolytics more efficacious than a single antidepressant for specific disorders and symptoms?
Is the combination of newer antidepressants with psychosocial therapies better than newer antidepressants alone for treating or maintaining remission for depressive disorders in adults?
Are newer pharmacotherapies plus augmenting agents (e.g. lithium, pindolol) more effective than pharmacotherapy alone for treating adults with depressive disorders?
Are newer antidepressants more effective than placebo, older agents, or psychosocial therapies for maintaining remission in adults with depressive disorders?
What common adverse effects of newer antidepressants have been identified in RCTs, and does their frequency vary significantly from one agent to another?
Do trials show varying adherence rates among newer antidepressants and between newer agents and older ones?
Do trials show varying rates between total drop-out, drop-out for adverse events and drop-outs for lack of efficacy?
What uncommon but serious adverse effects of newer antidepressants have been reported, and what is their frequency?
The main findings of this report, which are summarised in the published journal article (see Other Publications of Related Interest), are as follows:
Newer antidepressants compared with placebo in major depression (81 trials): overall, 51% of patients assigned to active treatment, compared with 32% of those assigned to placebo, experienced at least a 50% improvement in depressive symptoms. Relative benefit (calculated from 47 trials) was 1.6 (95% CI: 1.5, 1.7).
Statistically-significant publication bias was discovered for these trials (p=0.002).
Newer antidepressants compared with placebo in dysthymia (5 trials): overall, 59% of patients assigned to active treatment, compared with 37% of those assigned to placebo, experienced at least a 50% improvement in depressive symptoms. Relative benefit (calculated from 4 trials) was 1.7 (95% CI: 1.3, 2.3).
Newer antidepressants appear to be effective in elderly patients and in primary care patients. Data are insufficient to determine the relative efficacy of newer antidepressants and placebo in subsyndromal depression, depression with coexisting medical or psychiatric illness, or depression in adolescents.
Newer antidepressants were found to be as effective as older ones in major depression (150 trials) and in dysthymia (5 studies).
No difference between the efficacy of newer antidepressants was found, although there was a trend for fluoxetine to be less efficacious than other individual selective serotonin re-uptake inhibitors..
Overall drop-out rates did not differ between any of the newer antidepressants. Adverse events are discussed.
Hypericum versus placebo in mixed depressive disorders (6 studies): overall, 62% of patients assigned to hypericum, compared with 38% of those assigned to placebo, experienced at least a 50% improvement in depressive symptoms. Relative benefit (calculated from 6 trials) was 1.9 (95% CI: 1.2, 2.8).
Hypericum versus tricyclic antidepressants in mixed depressive disorders (5 studies): overall, 62% of patients assigned to hypericum, compared with 61% of those assigned to tricyclic antidepressants, experienced at least a 50% improvement in depressive symptoms. Relative benefit (calculated from 6 trials) was 1.2 (95% CI: 1.0, 1.4). Publication bias was identified (p=0.009)