Dropout from randomised controlled trials of psychological treatments for depression in children and youth: a systematic review and meta-analyses

Background: Depression is a prevalent and disabling condition in youth. Treatment efficacy has been demonstrated for several therapeutic modalities. Acceptability of treatments is also important to explore and was addressed by investigating treatment dropout using meta-analyses. Methods: A systematic search was conducted using MEDLINE, CINAHL and PsycARTICLES databases. Peer reviewed randomised controlled trials investigating psychotherapy treatment of depression in children and youth (aged up to and including 18 years) were included. Proportion meta-analyses were used to calculate estimated dropout rates; odds ratios assessed whether there was greater dropout from intervention or control arms and meta-regressions investigated for associations between dropout, study and treatment characteristics. Results: Thirty-seven studies were included (N=4343). Overall estimate of dropout from active interventions was 14.6% (95% CI 12.0-17.4%). Dropout was equally likely from intervention and control conditions, aside from family/dyadic interventions (where dropout was more likely from control arms). There was some suggestion that interventions offering more sessions and longer duration had less dropout and of less dropout from IPT than other interventions. There were no significant associations between dropout and study quality, CBT, family or individual versus other approaches. Limitations: Lack of consistent reporting decreased the factors which could be analysed. Conclusions: Dropout from depression treatment in children and youth was similar across different types of intervention and control conditions. Future treatment trials should specify minimum treatment dose, define dropout and provide information about participants who dropout. This may inform treatment choice and modification of treatments.