BACKGROUND: Reducing adolescent smoking is critical for improving long-term health and economic outcomes. However, many interventions inadvertently exacerbate socio-economic inequality. To address this, we propose a network-based social contagion model to assess innovative smoking cessation interventions among adolescents that can reduce both prevalence and inequalities.
METHODS: Our model describes how students influence each other’s smoking behaviour through social interaction. Each student transitions between three states: never-smoker, smoker, or quitter. Smoking cessation interventions are delivered via counselling sessions, modelled as external influences similar to social contagion.
To quantify intervention impact, we used Bayesian calibration (BOLFI) and real-world data from American students in grades 9 to 12. We simulated various strategies for targeting students (e.g., by social connectivity) and allocating intervention resources (e.g., contact hours) across the smoking population. The reduction in prevalence after 12 months was compared to random interventions.
RESULTS: Two strategies emerged as particularly effective. First, spreading resources across more than 40% of smoking students provides no additional benefit due to diminishing returns. Instead, focusing on fewer students can achieve the same impact while reducing operational costs. Second, prioritising smokers with fewer social connections outperforms targeting highly connected smokers. This improves intervention effectiveness and disproportionately benefits adolescents from deprived backgrounds, who tend to have fewer connections.
Combining these two strategies boosts cessation rates by an additional 1% compared to standard counselling: a modest but meaningful improvement, with the added benefits of lowering costs and addressing inequalities.
CONCLUSIONS: Our results suggest that targeting students with fewer connections, up to 40% of the smoking population, simultaneously outperforms random targeting, improves equity, and reduces costs, which makes it a compelling strategy for policymakers seeking to maximise impact.
Also, although this study focuses on adolescents, our framework can be adapted to other contexts, offering researchers a tool for designing effective cessation strategies with a broader reach.