What if nobody smoked? - attributable mortality estimates using system dynamics modeling

Abstract Background Smoking remains one of the most significant and modifiable risk factors contributing to preventable mortality worldwide. This study applies a system dynamics modeling approach to estimate smoking-attributable mortality under a hypothetical scenario where smoking is entirely eliminated in the Belgian adult population, providing a comprehensive forecast of associated health gains over time. Methods We developed a system dynamics model simulating the Belgian population, disaggregated by age, sex and the sub-national region, using the deSolve R package. The model calculates smoking initiation, relapse and cessation rates based on smoking trends projected with Bayesian modeling from the Belgian National Burden of Disease (BeBOD) study, which incorporates historical cigarette smoking trends from the Belgian Health Interview Survey. A counterfactual scenario is constructed assuming a complete elimination of smoking exposure. The primary outcome is all-cause mortality attributable to smoking between 2000 and 2040. Preliminary results The system dynamics model estimates that in 2025, approximately 11,300 deaths or 15% of all deaths in Belgium could be attributable to smoking in Belgium. The model predicts a steady decline in smoking-related deaths from 2000 through 2040, with a slower rate of decline observed towards the end of the period. Under the smoking elimination scenario, between today and 2040, over 175,000 smoking-attributable deaths could be prevented, or an average of nearly 11,000 per year. Conclusions System dynamics modeling highlights the significant mortality benefits of smoking eradication and provides valuable insights for health policy planning in Belgium. The sex, region and age-specific estimates further highlight demographic disparities in health gain, which could inform targeted intervention strategies. This dynamic modeling approach offers a context-specific and time-sensitive tool for quantifying the impact of potential tobacco control policies. Key messages • System dynamics modeling reveals substantial mortality reduction following smoking elimination. • Realistic and dynamic models of smoking behavior can inform the design and estimate long-term health effects of health policies.