In the review, only fluoxetine was found to be statistically significantly more effective than placebo; and the researchers suggest that it is ‘probably’ the best option to consider when a pharmacological treatment is indicated.
A commentary notes that these findings have disturbing implications for clinical practice. It suggests that there needs to be a conscious correction for favourable misrepresentation of outcomes in published and unpublished study reports, and a reduction should be applied to the reported benefit of a drug, while routinely assuming that its harms are more serious and frequent than reported. Only if the discounted benefit outweighs the boosted harm should the treatment be prescribed, though for antidepressants in adolescents, it acknowledges that this equation will rarely favour prescribing, and almost never in younger children. It notes that opposing this approach will be most psychiatrists and many GPs who have vast experience of prescribing antidepressants to adolescents, and many will believe that their clinical experience overrides any scepticism introduced by this research. The commentators accept this seems justified but providing they are honest with themselves and their patients that such prescribing is unsupported by evidence from RCTs. In addition, a second clinical implication is a need for an alternative to the guideline recommendation of prescribing antidepressants when evidence-based psychotherapy is unavailable. The commentary highlights that clinicians should not be pressured into prescribing just because they have no capacity to offer evidence-based psychotherapy. Furthermore, the effect of misreporting is that antidepressants, possibly including fluoxetine, are likely to be more dangerous and less effective treatments than has been previously recognised, so there is little reason to think that any antidepressant is better than nothing for young people. One final message from this study is about how data are managed, as the researchers reported that they could not be confident in the accuracy of published information used for their analysis because they were unable to access individual patient-level data. The commentary notes that patients who take part in RCTs have a right to expect that maximum benefit will come from the data they generate.