Building National-Scale Clinical AI Communities of Practice

Presentation Time: 08:30 AM - 10:00 AM


There is an ongoing narrative in policy and practice that AI in healthcare represents an ever-changing and
difficult-to-understand “Wild West.” This narrative can incite fear of AI and, as a result, hamper the effective design,
implementation, and regulation of tools that can and should improve health and healthcare delivery. Additionally,
this narrative creates confusion as to genuine issues surrounding the design and implementation of AI in healthcare,
including but not limited to navigating constantly evolving technical capabilities, monitoring for and addressing the
entrenchment of systemic data biases, and mitigating the potential for increasing disparities in equitable delivery of
care as a function of access to modern AI. There are many examples of what good AI looks like, how AI can be used
to improve patient and provider experience, and the cost, quality, safety, and outcomes of care delivery. Such examples
must be illuminated to demonstrate how appropriate design processes, governance, and operating models within
healthcare systems can be studied to create effective regulation and identify gaps in knowledge and practice that must
be addressed to help realize AI’s full potential in the healthcare domain. Central to achieving such goals is forming
communities of practice (CoPs) capable of convening diverse stakeholders, facilitating the exchange of knowledge,
and advocating for relevant research, practice, and policy agendas. This panel will explore the current state of such
CoPs and the unique and critical role of the biomedical informatics community in such a context.

Speaker(s):
Philip Payne, PhD, FACMI, FAMIA
Washington University in St. Louis, Institute for Informatics, Data Science, and Biostatistics (I2DB)

Nigam Shah, MBBS
Stanford University

Julia Adler-Milstein, PhD
UCSF School of Medicine

Mark Sendak, MD, MPP
Duke Institute for Health Innovation

Peter Embi, MD
VUMC