Depression and associated anxiety in primary care: A double-blind comparison of paroxetine and amitriptyline

Abstract

Background: A meta-analysis of the efficacy and tolerability of selective serotonin reuptake inhibitors (SSRIs) against tricyclic antidepressants (TCAs) in depressed patients was carried out. Methods: Efficacy data from 102 randomised controlled trials (10 706 patients) were pooled to provide a summary variance-weighted effect size. Tolerability data from 95 studies (10 553 patients) were combined to give variance-weighted relative risk of drop out for all reasons and for adverse effects from each study. The effect of age, treatment setting, severity and TCA dose were examined as well as the performance of individual SSRIs and TCAs where there were sufficient studies. Results: There is no overall difference in efficacy between SSRIs and TCAs (effect size −0.03, 95% confidence interval −0.09 to 0.03). TCAs do appear more effective in in-patients (−0.23, −0.40 to −0.05) and amitriptyline is more effective than SSRI comparators (−0.14, −0.25 to −0.03) but publication bias cannot be excluded. The SSRIs are better tolerated, with significantly lower rates of treatment discontinuations overall (relative risk 0.88, 0.83 to 0.93; number needed to treat 26) and due to side effects (0.73, 0.67 to 0.80; number needed to treat 33). Individual SSRIs show a similar advantage except for fluvoxamine which does not differ from the TCAs. Individual TCAs show a similar disadvantage in tolerability compared to SSRIs except for dothiepin against which SSRI treatment results in more side-effect related drop outs (2.64, 1.50 to 4.63; number needed to harm 12). Limitations: The evidence is from short-term studies and subgroup analyses may result in chance results. Conclusions: Overall efficacy between the two classes is comparable but SSRIs are not proven to be as effective as TCAs in in-patients and against amitriptyline. SSRIs have a modest advantage in terms of tolerability against most TCAs.