Equitable Diagnostics: What Laboratory Medicine Can Learn From Algorithmic Fairness

This lecture offers the following credit types: CME, P.A.C.E.®
 

As laboratory medicine moves toward greater integration of AI-driven tools, we must recognize the tendency of these algorithms to produce biased outputs, potentially exacerbating existing healthcare disparities. The field of algorithmic fairness provides a rigorous framework for defining fairness, identifying bias mechanisms, and engineering strategies to mitigate inequities in AI models. However, these lessons are not just relevant to the future of AI-based laboratory diagnostics—they also apply to our current laboratory practices, where bias arising throughout the total laboratory testing process can lead to inequitable patient outcomes. By drawing on insights from algorithmic fairness, laboratory medicine can proactively address inequities in both present-day workflows and future AI-driven decision-making, ensuring more just and effective patient care.

Originally presented at the Rocky Mountain Section ADLM Spring Seminar during April 2025.

Lecture Presenter

Mark A. Zaydman, MD, PhD

Mark A. Zaydman, MD, PhD

Assistant Professor of Pathology and Immunology
Washington University School of Medicine

Mark A. Zaydman, MD, PhD, is an assistant professor of Pathology and Immunology at Washington University School of Medicine in St. Louis, Missouri. His work helps clinical laboratory professionals improve the quality, safety, and value of laboratory testing by enabling them to leverage healthcare data and data analytics to inform operational and clinical decisions so that patients can enjoy better healthcare at reduced costs to the patient and their institutions. He serves on the ADLM Data Analytics Steering Committee.

Objectives

After this presentation, participants will be able to:

  • Define metrics of algorithmic fairness
  • Describe the different ways that AI models incorporate bias
  • Discuss applying the framework of algorithmic bias to ensure fairness as a quality domain in current laboratory practice

Sponsored by:

Spencer Fox Eccles School of Medicine at the University of Utah, Department of Pathology,
and ARUP Laboratories