Machine Learning and Structural Bias in Health Care

New ebook from Center for Connected Medicine captures key discussion points from senior leaders in health care data science and artificial intelligence

In November 2020, the Center for Connected Medicine (CCM) convened a group of experts in health care data science, artificial intelligence (AI), and machine learning (ML) for a virtual roundtable on the important and timely topic of bias in data and health disparities.

The invitation-only Top of Mind Exchange: Artificial Intelligence virtual roundtable was moderated by Robert M. Califf, MD, Head of Clinical Policy and Strategy for Verily and Google Health, and included experts from academia, government, health systems, and industry.

Eleven experts participated in the virtual roundtable, including Derek Angus, MD, UPMC’s Chief Healthcare Innovation Officer; Pamela Peele, PhD, Chief Analytics Officer of UPMC Health Plan and UPMC Enterprises; and Yubin Kim, PhD, Technology Director of UPMC Enterprises. These UPMC leaders spoke with experts from MIT, the National Institutes of Health, Harvard, Johns Hopkins, and other institutions.

The CCM’s report from the event, “Machine Learning and Structural Bias in Health Care,” captures high-level summaries of the discussion.

The discussion acknowledged the excitement and potential for AI/ML to benefit health care while also calling on leaders to confront challenges related to bias in data and the broader health care system.

Against the backdrop of social justice issues receiving significant attention during 2020, the roundtable discussion took on even greater significance. As roundtable participants pointed out, not only are there issues with bias in algorithms but there are also serious problems with disparities in health and outcomes across the medical industry.

A pivotal question for leaders and innovators to consider is, will advanced technology be a force for correcting or perpetuating long-standing inequities in society generally and medicine specifically? The report captures eight next steps and points of importance for tackling bias in data and structural inequities in health care.

Learn more by downloading “Machine Learning and Structural Bias in Health Care” from the CCM website.