Project 4 will evaluate the benefits, harms, and costs of personalized screening using one of the CISNET models of breast cancer. We will evaluate the current approaches to personalized screening in WISDOM and optimize the risk thresholds for the new subtype-specific risk models developed in this grant (Projects 2, 3) at the population level in the US. We will compare personalized screening to the screening recommendations of 3 guidelines widely used in the US (ACR, ACS, USPSTF). Given that the natural history of breast cancer is long, it is not possible to perform randomized trials comparing multiple screening strategies to assess their impact on breast cancer morbidity and mortality. The MISCAN-Fadia is one of six CISNET models of breast cancer in the US refined over 20 years of NCI funding. It allows us to estimate and compare the impact of different approaches to breast cancer screening on long-term breast cancer outcomes to complement the evidence generated from randomized trials such as WISDOM. Modeling has the advantage of combining evidence from multiple high-quality data sources and simulating the impact of screening strategies on both harms and benefits in subgroups of women based on their age, breast density, comorbidity level, race, and risk factor profile. The results of simulation modeling provide critical data that have been used to guide national screening recommendations globally. Most recently, the MISCAN-Fadia model was one of the CISNET models used to guide the 2023 draft United States Preventative Services Task Force (USPSTF) breast cancer screening guidelines. The aims of the project will implement the personalized screening approaches used in WISDOM 1.0 and 2.0 in the MISCAN model to project their impact on the benefits, harms and costs of breast cancer screening and treatment and compare these outcomes to those from other screening strategies. We will then incorporate the risk prediction tools for fast and slow growing cancers developed in the other projects in the grant into the MISCAN model in order to identify the optimal risk thresholds to use for the updated risk prediction tools. For example, we will systematically vary the risk thresholds for fast growing tumors used to recommend more intensive screening (MRI alternating with mammography every 6 months) and the risk thresholds for slow growing tumors used to recommend endocrine therapy and less frequent screening to estimate their impact on breast cancer specific mortality and other outcomes. Scenarios that result in a gain of at least 5 quality adjusted life years (approximately the gain from the move to tomosynthesis from digital mammography) with no increase in total costs will be presented to a multidisciplinary group of stakeholders to allow them to choose the thresholds that will be implemented in the next iteration of risk-based personalized screening in the WISDOM study.