The OncoSim Cancer Simulation Platform: A Tool to Project the Population Effects of Cancer Control Interventions in Canada

Abstract

Background: Cancer control interventions exert their effects over multiple decades. To evaluate diverse and competing opportunities to reduce future cancer burden it is desirable to understand long-term effects prior to any new program implementation or significant change. Internationally, modeling is becoming an accepted source of planning information for decision-makers. Aim: We will describe the construction and use of the OncoSim microsimulation model, which was developed to evaluate cancer control strategies in Canada. Methods: OncoSim is a suite of models (cancers of the lung, colorectum, cervix and breast, plus a composite 32-cancer model) used to address key policy questions and support decision-making. It is led by the Canadian Partnership Against Cancer with model development by Statistics Canada. OncoSim incorporates risk factors, cancer natural history, screening, treatment, survival and end-of-life care. Wherever possible it is informed by Canadian data sources. Models are calibrated to reproduce a range of cancer-specific statistics, e.g., current and historical Canadian cancer-specific incidence and mortality, smoking patterns, and results of screening. The site-specific models have undergone further validation by replicating reported short-term effects of cancer prevention and screening interventions. Users may customize interventions through modifying input parameters. Outputs include incidence, mortality, costs, cost-effectiveness, and resource utilization. Users from the public sector have access at no cost to OncoSim and receive extensive support from a multidisciplinary technical team. The model is continually updated to incorporate emerging knowledge. Results: OncoSim has been used to support cancer control decision-making at the national and provincial/territorial levels. Applications include: national guidelines recommendations for colorectal and lung cancer screening; comparison of cytology vs. HPV based cervical cancer screening; and integration of smoking cessation into low-dose CT lung cancer screening. Conclusion: Validated simulation models such as OncoSim can be a versatile and efficient tool for cancer control planners to evaluate and prioritize cancer control strategies.