Is high-quality of trials associated with lower treatment efficacy? A meta-analysis on the association between study quality and effect sizes of psychological interventions for pediatric PTSD.
children and adolescents with PTSD
psychological interventions
waitlist, active control conditions
effect size
Abstract
This meta-analysis aimed at providing an up-to-date estimate on the efficacy of psychological interventions for pediatric PTSD and to analyze the association between treatment efficacy and study quality. We systematically searched PsycINFO, Medline and recent meta-analyses for randomized controlled trials (RCTs). RCTs were eligible if a) they included at least 10 participants per group, b) compared a psychological intervention to a control condition or another psychological intervention and c) mean age was below 19 years. Study quality was assessed independently by both authors on the basis of eight quality criteria. We explored the potential associations between study quality and effect sizes in three ways. Firstly, we compared effect-sizes of high-quality vs. lower-quality studies. Secondly, we analyzed study quality as a continuous predictor of effect sizes. And thirdly, we examined the relationship between the eight individual quality criteria and effect sizes. A total of 46 eligible RCTs were included in the meta-analysis. Psychological interventions produced a large effect size when compared to waitlist (g = 1.07, k = 23, NNT = 1.81) and a medium effect size when compared to active control conditions (g = 0.60, k = 15, NNT = 3.03) at post-treatment. Overall, study quality was moderate. Comparisons of high-quality trials (k = 16) with lower-quality trials (k = 30) produced only non-significant findings in main-analyses as well as moderator sub-analyses. Study quality as a continuous variable was also not found to be related to effect sizes in any of the main analyses, nor was any of the eight individual quality criteria. The summary of the available literature strongly suggests that psychological interventions are effective in treating PTSD in children and adolescents. No significant associations between study quality and treatment efficacy were observed.