As I’ve written on this website previously, the prescription of antidepressants for children and adolescents is controversial, with warnings being issued relating to a possible paradoxical increase in suicidal thought and behaviour following prescription. However, experiences of low mood and depression for young people remain common, affecting approximately 1 in 20 adolescents at any given time.

Most clinical guidelines now recommend psychological therapies as the first point of treatment for depression in children, but waiting lists for therapies are often long or in some cases they are simply not available. The demand to provide urgent support and relief of distress is strong and understandable, therefore antidepressant drugs can often seem like a good option to reach for. Are we justified in prescribing such treatment? If so, which of the many drugs available should be offered?

Aims

This question of antidepressant superiority was addressed in a recent network meta-analysis reported by Cipriani and colleagues in which they sought to:

…comprehensively compare and rank antidepressants for the acute treatment of major depressive disorder in children and adolescents.

Methods

Network meta-analysis represents a powerful inferential statistical approach whereby the findings from previous clinical trials can be combined together in a comparative manner to generate cross comparison between various treatments. In this way it is possible to generate ‘hierarchies’ of treatments, on the basis of available evidence, relating to outcomes such as treatment response, or adverse event rates.

The authors therefore began their study by identifying pertinent literature relating to the efficacy of various antidepressant drugs for the treatment of depression in childhood and adolescence (e.g. Fluoxetine, Venlafaxine, Mirtazapine, Citalopram etc).

Outcome measures

For their primary outcome measures the authors considered:

1. Change in depressive symptoms (on a validated measure for depression in children and adolescents)
2. Discontinued treatment owing to any adverse events

Secondary outcome measures considered included:

1. Response rates (either proportion of the trial population achieving a 50% reduction in symptoms, or a change of ‘much’ or ‘very much’ improved on the Clinical Global Impression Scale)
2. All cause discontinuation
3. Suicidal behaviour or ideation

Statistical analysis

Outcomes were expressed either as standardised mean difference (continuous outcomes, e.g. symptom severity) or odds ratios (categorical outcomes, e.g. discontinued treatment). 95% credible intervals were then calculated for each measure. Variation between outcomes was assessed using a measure of heterogeneity (the difference between varying trial findings) and finally the presence of publication bias was appraised using a visual inspection of trial findings within a funnel-plot.

Outcome measures were then compared through a network meta-analysis process to generate a hierarchical comparison of the drugs.

Results

• The authors identified 31 publications describing 34 randomised controlled trials of 14 different antidepressant drugs
• These trials considered a total of 3,106 participants receiving antidepressants vs 2,154 receiving placebo
• Approximately half of trial participants were female (53%) and the average age was 13.6 years
• The median duration of acute treatment was 8 weeks
• In terms of methodological quality, most identified trials were graded as being of low, or very low quality, and at high risk of bias
• The I² heterogeneity scores were 33·21% for efficacy and 0% for tolerability (I² values are low – 25%, moderate – 50% and high – 75%).

Antidepressant efficacy and adverse events

• Of all the reviewed antidepressants, only Fluoxetine achieved statistically significant superiority over placebo in terms of efficacy. This finding was associated with a wide credible interval (standardised mean difference -0·51, 95% credible interval [CrI] -0·99 to -0·03), that came close to the point of no-effect (0.0) raising concerns over the clinical significance of the observation
• In terms of hierarchy therefore, Fluoxetine was rated as superior to other antidepressants in terms of efficacy and tolerability
• Venlafaxine was associated with a statistically significant increase in suicidal ideation and behaviour in comparison with placebo and other antidepressants.

Discussion

The authors conclude:

When considering the risk-benefit profile of antidepressants in the acute treatment of major depressive disorder, these drugs do not seem to offer a clear advantage for children and adolescents. Fluoxetine is probably the best option to consider when a pharmacological treatment is indicated.

Implications

As an accompanying editorial (Jureidini, 2016) to this research states the findings have:

…disturbing implications for clinical practice, concluding as it does that the risk-benefit profile of antidepressants in the acute treatment of depression does ‘not seem to offer a clear advantage for children and adolescents.’

Both the authors of the main review, and the accompanying editorial, suggest that even the findings relating to Fluoxetine should be interpreted with caution as the clinical significance is unclear, and other, unmeasured, factors may be present that favour the medication outcome within poor quality randomised control trials.

This review therefore implies that any prescribing of antidepressants for young people must be conducted in a particularly careful manner; with close attention to psychosocial supporting work alongside the prescription and awareness for the development of any adverse effects or emergent suicidal ideation.

Links

Primary paper

Cipriani, A., Zhou, X., Del Giovane, C., Hetrick, S. E., Qin, B., Whittington, C., et al. (2016). Comparative efficacy and tolerability of antidepressants for major depressive disorder in children and adolescents: a network meta-analysis. Lancet, 388(10047), 881–890. http://doi.org/10.1016/S0140-6736(16)30385-3

Other references

Jureidini, J. (2016). Antidepressants fail, but no cause for therapeutic gloom. Lancet, 388(10047), 844–845. http://doi.org/10.1016/S0140-6736(16)30585-2

Photo credits

• Random Literature Council CC BY 2.0
• Hernán Piñera CC BY 2.0
• Mike Hartnett CC BY 2.0