Can fluoxetine's diminished efficacy in pediatric depression be explained by study sites, baseline severity, age, and psychological interventions? A secondary exploratory meta-regression.

BACKGROUND: Fluoxetine's efficacy estimates in pediatric depression trials have declined to the range of placebo equivalence. We investigated whether changes in trial characteristics (number of study sites (NOSS), age, baseline severity, or added psychological interventions) could explain this trend. METHODS: We conducted a secondary exploratory analysis based on our recent meta-analyses of pediatric fluoxetine depression trials. The outcome was symptom change on the Children's Depression Rating Scale Revised (CDRS-R). We used meta-regression to explore the association between NOSS, age, depression severity, and the addition of psychological interventions with efficacy and placebo response. RESULTS: NOSS increased over time and was not significantly associated with outcomes. When log-transformed, NOSS was associated with decreased efficacy (B = 1.57, 95% CI 0.24 to 2.90, p = 0.02) and greater placebo response (B = -1.90, 95% CI -3.41 to -0.40, p = 0.01). However, these associations were driven by an outlying single-center trial rated as high risk of bias and did not survive p-value adjustment. Age, baseline severity, and psychological interventions were not significantly associated with efficacy or placebo response. CONCLUSION: NOSS increased over time but the apparent associations with decreased efficacy or increased placebo response were driven by an outlying trial at risk of bias. Age of participants, baseline severity, or added psychological interventions also failed to explain fluoxetine's diminished efficacy. Limitations include the potential ecological fallacy in analysis of age and baseline severity. Overall, these findings do not support the assumption that fluoxetine's meta-analytic efficacy estimates are substantially influenced by these trial characteristics.