Disparities in Prioritization and Admission Decisions in Emergency Departments

Emergency departments (EDs) in the U.S. and around the world face numerous challenges. Of particular note is understanding and addressing demographic disparities associated with various decisions in an effort to provide equitable treatment to all patients. In this dissertation, we first investigate the rooming prioritization decision by examining and comparing timestamps within electronic health record (EHR) data by means of weighted logistic regression and matched 1:M case-control logistic regression. Our results show that demographic factors remain significant even after accounting for differences in patient and ED operational conditions. We then examine the disposition decision and determine which patient characteristics are more important as part of that decision with random forests and logistic regression. Finally, using a priority queueing system we investigate the impact of erroneous triage decisions in emergency departments from both operational efficiency and disparity points of view. We analyze several different priority policies and find that, particularly in crowded systems, some of these policies can be used to keep disparities mostly under control without sacrificing significant operational efficiency.