Background Childhood mental health issues have long-lasting effects on adult health, and economic outcomes. The Incredible Years Teacher® Classroom Management (IY-TCM) intervention is associated with short-term improvements in reducing difficulties. The aim of this study is to estimate the long-term impact and cost-benefit of the IY-TCM intervention compared to no intervention.

Methods Individual participant data from the STARs trial (2075 children aged 4-9 years with a 30-month follow-up) was re-analysed using statistical regression to examine the impact of the IY-TCM intervention on the conduct and impact sub scores of the strengths and difficulties questionnaire (SDQ). LifeSim, a microsimulation model, was used to estimate the long-term effects of the intervention by linking the short-term SDQ changes to adolescent and adult health and economic outcomes using various data sources including longitudinal studies (e.g., Millenium cohort study), surveys, administrative records, and existing research on the relationship between early life circumstances and long-term outcomes. The LifeSim model was adapted to incorporate both teacher- and parent-reported SDQ and account for individual heterogeneity. Cost-benefit analyses were conducted with subgroup analyses based on deprivation, conduct scores and parental depression in the simulated baseline population. Benefits were measured in quality adjusted life years and wellbeing adjusted life years. All analyses were conducted in R.

Results Regression analyses show significant predictor variables for intervention effectiveness including interaction effects between deprivation and baseline SDQ. LifeSim results indicate small gains in long-term outcomes including mental health, and cost-benefit analyses estimated that the IY-TCM intervention is cost-effective but with a large amount of uncertainty (Net monetary benefit (NMB)=£21, Estimated CI = -£195, £240). Benefits are greater, with greater certainty of cost-effectiveness, for some subgroups such as those with high deprivation (NMB = £231, Estimation CI - -£99, £231) and high conduct scores at baseline (NMB=£315, Estimated CI = £45, £483). For a typical year size group in England (n = 775000), the total net social benefit was estimated to be £111m.

Discussion IY-TCM is estimated to be cost-effective compared to no intervention, but with a large amount of uncertainty around costs and benefits. The LifeSim model focuses on the impacts the conduct SDQ sub score and greater impacts may have been estimated if modelling changes in all SDQ subscales. Greater benefits for pupils with difficulties at baseline suggest that the intervention may be more cost-effective for schools in more deprived areas with a greater proportion of children with conduct problems.