Center for Bioethics & Health Law Colloquium

Jamie Zelazny, PhD, MPH, RN
Assistant Professor of Health & Community Systems
School of Nursing, University of Pittsburgh

Abstract: This work-in-progress discussion will explore the ethical challenges that must be considered when designing and conducting research studies involving the use and implementation of suicide risk prediction algorithms. Natural language processing (NLP) and machine learning (ML) provide cutting edge tools to mine large data sets for the development of predictive algorithms. The application of these methods to electronic health record (EHR) information and social media (SM) data can potentially aid in developing algorithms able to predict risk for psychiatric illness and suicidality. Careful consideration of the ethical issues involved in the development and implementation of these methods is needed. Ethical concerns are inherent with the use of risk algorithms for the prediction of illness in general. These concerns are magnified in vulnerable populations, especially in those with mental health disorders or suicidality.