Drug dose as mediator of treatment effect in antidepressant drug trials: the case of fluoxetine.

Abstract

This study aimed at investigating whether dose is a mediator of treatment effect in fluoxetine-randomized trials. Specifically, we investigated whether dose was higher in trials in which the aim was to demonstrate fluoxetine efficacy in comparison with older antidepressants and lower in trials in which the aim was to demonstrate a new drug's efficacy against fluoxetine. We applied the model developed by Baron and Kenny to investigate the mediational role of drug dose on treatment effect. We included all randomized controlled trials comparing fluoxetine with other antidepressants as monotherapy in the acute-phase treatment of unipolar major depression. A total of 173 studies were included. In 76 comparisons (44%), fluoxetine was the experimental antidepressant. A metaregression analysis indicated that after adjusting for possible confounders, studies where fluoxetine was the experimental agent were associated with a significant advantage for fluoxetine. However, the Baron and Kenny model revealed no mediational role of drug dose in influencing treatment effect. The outcome of fluoxetine-randomized trials changed according to whether this drug was used as a new compound or as a reference. This finding cannot be attributed to antidepressants dose, as dose failed to emerge as mediator of treatment effect in the Baron and Kenny approach.