Risk and disease prediction tools are common in clinical practice. They are vital to high-quality care—simplifying complexity in evidence-based fashion. Yet, despite being used everyday use, many tools are now recognized to be racially biased, skewing clinical judgments, to contribute to worse outcomes for minoritized populations. This session brings together experts in health equity and clinical/prediction testing to discuss this manifestation of structural racism in pulmonary and critical care medicine. Attendees will learn about real world examples of biased tools used every day in our field, build a framework for understanding how biased tools emerge, and review mitigation strategies.
• Describe new findings about current biases in care delivery relevant to pulmonary/critical care medicine
• Identify the limitations of current approaches to clinical decision-making tools and treatment recommendations and evidence derived from machine learning approaches
• Apply expert-consensus mitigation strategies for bias to improve the quality of clinical care
Aaron Baugh, MD
Jennifer Taylor-Cousar, MD, MSc, ATSF
Aaron Baugh, MD
Thomas Valley, MD, MSc
Deepshikha Ashana, MD, MBA, MS
Nichole Tanner, MD, MSCR
Lori Perine, MBA
Deepshikha Ashana, MD, MBA, MS
Thomas Valley, MD, MSc
Introduction
Shortcomings of Standard Cystic Fibrosis Screening Tests Among Individuals of African Ancestry
Color-Blind or Context Blind? Understanding Race in Pulmonary Function Testing
Hold Your Breath: The Journey to an Equitable Pulse Oximeter
When Technology Imitates Life: How Bias Manifests in Risk Prediction and Artificial Intelligence
Unequal Risk: Modifying Lung Cancer Screening Guidelines
The Bias of Machines
Summary
Discussion