This module is based on an extensive literature review of artificial intelligence and ethics, and algorthmic bias that was completed in Fall 2021. Relevant papers were analyzed for concepts of algorithmic bias and those concepts were placed into related themes. The initial review revealed 86 on topic articles from the previous 5 years with 886 instances of 141 concepts across 8 themes. To inform the development of a training module, the 5 most common concepts were compiled. Additional references were used as needed to fill in gaps. This module was created as part of a Capstone Project for an M.S. degree in biomedical informatics at OHSU. The module will have study period to collect and analyze anonymous feedback to understand how well the methodology contributed to the development of a training module on a topic for which there is no clear expert (pending IRB approval). The study will include collection of anonymous responses to the pre-module and post-module surveys in order to inform practices for training clinicians on artificial intelligence in health care. Comments from the Disqus pages for each of the case questions will not be used in this study. Additionally, any feedback provided via email will not be used. Participation by submitting the RedCap survey forms is encouraged but it is not a mandatory part of the module.
[click here for PDF of IRB protocol](pending IRB approval)
Benjamin Collins was a Clinical Informatics Fellow at Oregon Health & Science University at the time of developing this training module. He has since transitioned to Vanderbilt University Medical Center where he is a hospitalist and postdoctoral research fellow studying the ethical, legal, and social issues of artificial intelligence in healthcare. He trained in Internal Medicine at Temple University Hospital. During residency, he also completed an M.A. in Urban Bioethics with his Master's Thesis on "A Theory of Sociotechnical Justice in Healthcare." After completion of the CI Fellowship, he will go into a postdoctoral fellowship at Vanderbilt University focused on Ethics, Legal, and Social Issues in AI for health care.
Email: AlgorithmicBiasModule@gmail.com