Presentation (CTI)

OBJECTIVES: To evaluate the cost-effectiveness of low-dose computed tomography (LDCT)-based lung cancer screening strategies compared with chest radiography (CXR) among high-risk individuals aged ≥55 years in Thailand.
METHODS: A cost-utility analysis compared three lung cancer screening strategies—chest radiography with artificial intelligence assistance (CXR-AI), LDCT alone, and CXR-AI followed by LDCT—with conventional CXR. A decision tree linked to a lifetime Markov model was developed from a societal perspective. Lifetime costs, quality-adjusted life-years (QALYs), and incremental cost-effectiveness ratios (ICERs) were estimated. Costs were estimated in 2025 values and reported in 2026 US dollars (USD; 1 USD = 32.59 THB). Costs and outcomes were discounted at 3% annually. Probabilistic sensitivity analysis (5,000 simulations) and scenario analyses were performed.
RESULTS: At a lung cancer prevalence of 1.2%, all screening strategies increased discounted QALYs but also increased lifetime costs compared with CXR. Mean discounted lifetime costs per person were USD 549, 661, 752, and 818 for CXR, CXR-AI, LDCT alone, and CXR-AI followed by LDCT, respectively. Incremental QALY gains were 0.0078, 0.0091, and 0.0114, with corresponding ICERs of USD 14,204, USD 22,353, and USD 23,603 per QALY gained. All ICERs exceeded the Thai willingness-to-pay threshold of USD 4,909 per QALY. Cost-effectiveness improved as lung cancer prevalence increased. LDCT became cost-effective at an estimated prevalence of 6.7%, whereas CXR-AI became marginally cost-effective at 9.1%. CXR-AI followed by LDCT was not cost-effective in any scenario.
CONCLUSIONS: LDCT-based lung cancer screening is unlikely to be cost-effective in the currently defined high-risk population in Thailand. However, targeting screening to very-high-risk individuals, particularly those with multiple first-degree relatives with lung cancer, may improve cost-effectiveness and represent a more efficient use of healthcare resources.