Accounting for Patient Learning in Iterative Physician Decision Support Systems
Due to the ever-increasing complexity of the situations faced by physicians, there is a need for decision support systems. There have been two assumptions for these models: that patients with similar medical diagnoses should be treated identically, and physician decisions are made rationally according to standard medical practices. However, we know that acknowledgment of patient uniqueness is required for an optimal outcome, and with each interaction, a physician has the opportunity to update their medical decisions based on different patient characteristics. We posit that socio-economic conditions also play a role in a physician’s decision-making process. Therefore, to make appropriate medical treatment suggestions, a DSS needs to adapt for both patient demographics and medical diagnoses. The goal of this work is to understand if physicians do adapt their decision-making strategies for different patient types, as this would allow us to create a more accurate data-driven model by accounting for prior patient encounters.