Lifetime trajectories of mental ill-health are established in childhood. It has been estimated that 10% of young people aged 5 to 16 have a clinically-diagnosed mental health disorder. For some young people, mental health problems continue into adulthood; half of adults with mental ill -health report their first symptoms having occurred before age 14. The Royal College of Psychiatrists state “Since the majority of lifetime mental illnesses develop before adulthood, prevention targeted at younger people can generate greater personal, social & economic benefits than intervention at any other time in the life course”.

The mental health of young people has been the focus of a range of preventative interventions. Preventative interventions for common mental health problems are typically complex with well-defined, multi-level components based on theory or a logic model. When commissioning, or designing, an intervention it is critical to understand which components/ mechanisms are essential in making the intervention successful (or not). Indeed, where local circumstances require modification of interventions, or cost-reduction, having identified important components would allow locally adaptive implementation. The identification of effective components or the “active ingredients” in public health interventions is a growing area of research and approaches exist for extracting details from published studies. The effect of intervention components (individually or in combination) can be modelled in meta-analysis using meta-regression methods. A network meta-analysis (NMA) is a form of meta-regression which enables the simultaneous comparison of multiple interventions in a single statistical model, whilst retaining the distinct identity of each intervention analysed. In NMA it is not necessary to 'lump' or conflate all relevant interventions to form a single comparator for comparison with an active intervention. Recent work focusing on intervention components as the unit of analysis for NMA has highlighted the importance and feasibility of NMA in public health, and how it can be useful to explore and minimise heterogeneity in evidence syntheses. A component-based NMA is ideally suited to synthesising preventative mental health interventions since we can incorporate the complexity of interventions whilst providing a coherent and quantitative assessment of effectiveness necessary for economic evaluation.

Despite the importance of the question of resource use, there are few published high-quality, trial-based economic evaluations of preventative mental health interventions for young people. Available economic studies suggest some evidence of economic benefits, particularly from early-years interventions. However, available studies have not used common or consistent endpoints and few have conducted sensitivity analyses to explore uncertainty in costs and effect inputs. In the absence of a large, multi-arm RCT comparing all interventions with long-term follow-up, probabilistic economic modelling offers an alternative to estimate cost-effectiveness. Whilst acknowledging the difficulties associated with economic modelling in public health, a recent review of economic evaluations has called for modelling studies to examine mid- to longer-term costs and benefits across different contexts, settings and whether “multi-level” interventions are cost-effective. To our knowledge, no published economic evaluation of mental ill-health prevention has included a NMA. Through identifying key components of effective mental health interventions our findings could be used in guiding the translation of ‘in research’ effectiveness into lower-cost (yet still effective) interventions for implementation in public health practice and identify priorities for future areas of research.