School-based interventions to prevent anxiety, depression and conduct disorders in children and young people:a systematic review, network meta-analysis and economic evaluation

Background: UK schools increasingly have to respond to anxiety, depression, and conduct disorder as key causes of morbidity in children and young people (CYP).Objective: To assess the comparative effectiveness of educational setting-based interventions for the prevention of anxiety, depression, and conduct disorder in CYP.Design: Systematic review, network meta-analysis (NMA), economic evaluation.Data sources: MEDLINE, EMBASE, PsycINFO and CENTRAL searches to 4 April 2018 and NHS EED on 22 May 2019 for economic evaluations. No language or date filters applied.Main Outcomes: Post-intervention self-reported anxiety, depression or conduct disorder symptoms.Review methods: We included randomised/quasi-randomised trials of universal or targeted interventions for prevention of anxiety, depression or conduct disorder in CYP aged 4–18 years. Interventions were based on Cognitive Behavioural Therapy (CBT), mindfulness/relaxation, CBT+Interpersonal Therapy (IPT), third-wave, exercise, biofeedback, behavioural, IPT, bias modification and occupational therapy. Screening was conducted independently by two reviewers. Data extraction was conducted by one reviewer and checked by a second. Intervention and component-level NMA were conducted in OpenBUGS. A review of the economic literature and a cost-consequence analysis were conducted.Results: 142 studies were included in the review, and 109 contributed to the NMA. 57/109 were at unclear risk of bias for random sequence generation and allocation concealment. Heterogeneity was moderate. In universal secondary school settings, mindfulness/relaxation interventions (Standardised Mean Difference (SMD) -0.65, [95%CrI -1.14 to -0.19]) and CBT (SMD -0.15 [95%CrI -0.34 to 0.04]) may be effective for anxiety. CBT interventions containing a psychoeducation component may be effective (SMD -0.30, [-0.59 to -0.01]) at preventing anxiety post-intervention. For anxiety in targeted secondary settings there was evidence exercise was effective (SMD -0.47, [95%CrI -0.86 to -0.09]). There was weak evidence in universal (SMD -0.07, [95%CrI -0.23 to 0.05]) and targeted (SMD -0.38, [95%CrI -0.84 to 0.07]) primary settings, that CBT may prevent anxiety. For preventing depression in universal secondary settings, there was weak evidence that CBT (SMD -0.04, [-0.16 to 0.07]) and CBT+IPT (SMD -0.18, [-0.46 to 0.08]) may be effective. Third wave (SMD -0.35, [-0.70 to 0.00]) and CBT interventions (SMD -0.11, [-0.28 to 0.05]) containing a psychoeducation component may be effective at preventing depression post-intervention. There was no evidence of intervention effectiveness in targeted secondary, targeted primary, or universal primary settings post-intervention. Results for university settings were unreliable due to inconsistency in the NMA. A narrative summary was reported for five conduct disorder prevention studies, all in primary school settings. None reported the primary outcome at post-intervention. The economic evidence review reported heterogeneous findings from six studies. Taking the perspective of a single school budget and based on CBT intervention costs in universal secondary settings, the cost-consequence analysis estimated an intervention cost of £43 per student.Limitations: The emphasis on disorder-specific prevention excluded broader mental health interventions and restricted the number of eligible conduct disorder prevention studies. Restricting to interventions based in the educational setting limited the number of eligible university-level interventions.Conclusions: There was weak evidence of the effectiveness of school-based, disorder-specific prevention interventions, although effects were modest and the evidence not robust. CBT-based interventions may be more effective if they include a psychoeducation component.Future work: Future trials for prevention of anxiety and depression should evaluate CBT with and without a psychoeducation component, and include mindfulness/relaxation or exercise comparators, with sufficient follow-up. Cost implications must be adequately measured.