Back

Time for Tobacco Elimination: Modelling smoking cessation strategies and lung cancer screening in Singapore

He, Y.; Jin, S.; Zhang, X.; Fong, K. I.; Wang, Y.; Tan, K. B.; Soo, R.; Lim, J. T.; Dickens, B.

2026-05-08 public and global health
10.64898/2026.05.06.26352560 medRxiv
Show abstract

BackgroundLung cancer remains a major public health burden, with poor survival largely driven by late-stage diagnosis. With declining and very low smoking prevalence in Singapore at 4.7% in 2024 among 18-29-year-olds, questions arise about future screening efficiency, eligibility criteria, and the impact of smoking cessation, including tobacco elimination. MethodsWe developed a large-scale microsimulation model calibrated to real-world data, generating individual life histories, smoking trajectories, and disease progression for Singapores 4.18 million residents to project smoking prevalence and lung cancer burden. We evaluated 271 low-dose computed tomography (LDCT) screening strategies (by age, gender, uptake, and frequency) under five tobacco control scenarios, from status quo to a complete smoking ban, between 2025 and 2050. FindingsUnder the status quo, all screening strategies were cost-effective relative to the 2024 GDP per capita threshold ([~]SGD 120,000). Among strategies with [≤]10% overdiagnosis, annual screening of eligible ever-smokers aged 50- 80 years was most life-saving, yielding 51,312 (95% uncertainty interval: 36,821-72,830) QALYs at a total cost of SGD 12.2 (9.7-16.1) billion. Adding an immediate smoking ban increased QALY gains by 2.8 (2.2-3.5) times while reducing the total cost by 23.3% (17.0%-30.0%). Extending eligibility to individuals with lower smoking exposure or a first-degree family history remained cost-effective. InterpretationsTobacco elimination yields substantial health and economic benefits, while well-designed risk-based LDCT screening of residual high-risk populations remains cost-effective, supporting a continued role for screening even in settings with declining smoking prevalence.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.