Lung cancer is the leading cause of cancer death in the U.S. The potentially enormous costs of a screening program and the large societal burden of disease from lung cancer are compelling reasons to evaluate the cost-effectiveness of screening protocols This project uses a computer model of lung cancer to predict the number of lung cancer deaths that could be prevented by advances in prevention, screening and treatment. Costs of these advances are also predicted, so that we can identify strategies to reduce lung cancer deaths that are also good values.

The MGH Lung Cancer Policy Model (LCPM) is a microsimulation model of lung cancer development, growth, progression, detection, and survival. Individuals can develop multiple lung cancers of different histological types. A 'true' disease stage is assigned based on the individual's simulated disease characteristics (tumor size, nodal involvement, distant spread) and updated every cycle. An observed disease stage is also assigned, based on the individual's true disease state and any results from diagnostic or staging tests. Detection of a pulmonary nodule suspicious for lung cancer may be prompted by symptoms or by incidental detection during a thoracic imaging examination for reasons unrelated to lung cancer (e.g., trauma) or on a screening examination (depending on the scenario). Benign pulmonary nodules are simulated because they are not always distinguishable from lung cancers on imaging exams and may prompt clinical workups, with all attendant risks and costs. Because the LCPM simulates details of clinical events including specific staging examinations and treatment modalities, the model can be used to compare a wide range of patient management strategies. The LCPM can simulate the full range of cancer controls, from prevention (smoking cessation) to screening to treatment. Please visit the technical appendix for the LCPM for more information.

Applications using the LCPM

• Helical CT screening for lung cancer
• Calibration methods for disease simulation models
• CISNET
• Detection of lung cancer risks with serum metabolomics
• Expanding the National Health Accounts
• Developing approaches for modeling genomic and proteomic profiles in lung cancer