Using Natural Language Processing as a Tool to Identify Outcome Disparities in Clinical Ethics Consultations

Abstract: A largely unexplored area of clinical bioethics is whether there is evidence of health disparities in data associated with clinical ethics consultation notes entered into electronic health records at major medical institutions. While the templates for these notes have improved over time, crucial data are still contained or even “trapped” as narrative text inside them. Natural Language Processing (NLP) is a branch of Artificial Intelligence that uses machines to understand human language. Natural Language Processing has already shown value as a tool to extract important data from narrative text within electronic health records. This Flash Presentation will share a comprehensive literature review as well as the initial steps taken by the research team at the Parkinson School of Health Sciences and Public Health at Loyola University Chicago to discover whether disparities exist in clinical ethics consultation data. If disparities do exist, the researchers plan to identify the disparities and categorize them according to ethical principles and concepts.