High participation by target populations is one of the key success factors of publicly funded and population-based cancer screening programs. Patient preferences are a major factor distinguishing preventive screening from treatment, specifically regarding participation and choice of facility. Here, we explore the effect of facility attributes on patients’ choices for cancer screening and propose a quantitative modeling framework to study the impact of various system- and facility-level interventions by incorporating the choice behaviors of heterogeneous patient populations. EPPM integrates an empirical study on patient preferences and a simulation-based optimization model that incorporates the preferences of multiple patient types into an M/G/s queuing network of facilities, each working with appointment. Based on our framework's application via an extensive case study on the Quebec Breast Cancer Screening Program in Montreal, we provide insights into improving participation in cancer screening networks. Specifically, we identify trade-offs that women make when choosing facilities with different attribute configurations and show the heterogeneity in patient preferences. We also find that facility-level interventions, such as providing free parking and improving staff manners, are more effective and less expensive than system-level interventions.