2122 - Human-Centered Systems Modeling for Mobile Radiation Oncology Implementation

A. T. Price^1,2, C. Kueny³, G. D. Hugo⁴, L. E. Henke¹, and C. Canfield³; ¹Department of Radiation Oncology, University Hospitals Seidman Cancer Center, Case Western Reserve University, Cleveland, OH, ²Missouri University of Science & Technology, Rolla, MO, ³Missouri University Science & Technology, Rolla, MO, ⁴Washington University School of Medicine in St. Louis, Department of Radiation Oncology, St. Louis, MO

Purpose/Objective(s): At present, rural residents are less likely to receive radiotherapy due in part to the high travel burden required for daily treatment visits to a distant facility. This burden could be reduced by bringing mobile radiation oncology (MRO) units installed on trucks to remote areas with high cancer rates. However, due to the lack of precedence in this approach, this could be a high-risk venture for a healthcare network. Therefore, we developed a digital twin Agent-Based Model (ABM) simulation using systems engineering, opinion dynamic theory, and targeted survey data to guide and simulate the implementation of an MRO. Materials/

Methods: ABMs are defined as having a bottom-up modeling approach where individual agents (such as patients or physicians) are governed by “rules-of-engagement.” In our ABM, agent interactions were defined by patient and physician preferences to utilize MRO from collected survey data. General public cancer population data was collected from public health resources and data for rate of radiation therapy and rate of referral within our network was captured to model the current state of our cancer network. The MRO linac would visit two separate geographic locations over a two-week period. In our ABM, each patient is initiated with an MRO preference and cancer type. A patient then interacts with physicians (such as a consult) and family where their intent to utilize MRO is updated via opinion dynamic models e.g., Continuous Opinion Discrete Action and the Relative Agreement Algorithm. The patient’s preference to pursue MRO is calculated via the Theory of Planned Behavior. Three separate methods of MRO implementation strategy were tested by influencing the agents within the model. A grass-roots approach (GRA) by improving patient attitude towards MRO, a community physician focus approach (CA) by improving physicians’ attitude towards MRO, and a main-hub physician focus (HA). Rates of MRO usage and impact on network patient volume were evaluated.
Results: In a 2-year period, the GRA treated an average of 106.3±21.5 patients. The CA treated an average of 115.9±22.6 patients and HA treated an average of 132.3±29.1 patients. Although numerically higher, HA was not statistically higher than the other approaches. To demonstrate patient benefit, HA reduced patient travel by an average of 15.2±11.6 miles. HA had the highest utilization rates for both patients and physicians whereas CA did not improve utilization rate for the main-hub physicians. This may be due to the low spoke-to-spoke interactions within our network which may limit influencing power. All intervention strategies resulted in a 23.9% average reduction in patient volume at a facility closer to the MRO.
Conclusion: An ABM was successfully developed that could test potential intervention strategies in high-risk scenarios for healthcare technology deployment. MRO has potential to reduce patient travel burden and increase access to care.