Note to non-members: Most of our content is available only to HDAA members. However, we have made our March 2023 town hall on ethical AI in healthcare accessible to all. Click here for more on this informative and timely proceeding. 

HDAA TOWN HALLS

In March 2020, HDAA responded to the COVID-19 crisis with a series of online discussions to assess the impact on healthcare data (and lack thereof), the immediate need for predictive models, health systems relationship with public health agencies, the impact on healthcare employees and on the profession of healthcare analytics.

We met weekly in the beginning, when data was sparse and tension high, evolving to biweekly and then monthly conversations as the pandemic continued. More than 100 HDAA members contributed to those conversations.

The COVID sessions evolved into HDAA's Town Hall series, periodic presentations or panel discussions on issues of the day that affect healthcare data and analytics. One recurring theme is data and analysis around the health equity issue. Another is the impact of a remote healthcare work force.

Think of the town halls as keynote addresses, often by one of our members, occasionally an outside speaker. 

Upcoming events

None scheduled. Check this page, and our member home page, for updates.

Recently completed events

October 2024 | Baxter IV fluids disruption  
Open discussion, moderated by Pete Ivers Johns Hopkins Medicine

July 2024 | The National Wastewater Surveillance System
Jeff Mercante CDC

March 2023: | Ethical artificial intelligence for healthcare 
Jon Handler OSF Healthcare
Marzyeh Ghassemi MIT
Dwight Barry Seattle Children's Hospital
Anne Schaefer Kaiser Permanente
Moderator: Bryan Nice University of Washington School of Public Health

Four distinguished panelists – a physician, a researcher, and two data scientists – walked HDAA members through the unique ethical issues posed by artificial intelligence in healthcare. The discussion featured broad agreement on addressing equity and security, although with nuanced differences. Among the interesting questions raised:

• How can a researcher test the value of an AI model with limited (or no) access to the underlying data and assumptions?
• What are the ethical concerns of NOT taking advantage of AI advances?
• What practical steps can a healthcare institution take to assess the ethical integrity of a model?

Most of our content is available only to HDAA members, but we are making this important and timely proceeding available to all. We encourage you to take a listen.

Further reading suggested by our panelists:

• Marzyeh
• Is Fairness Only Metric Deep? Paper by Marzyeh and others
• Jon
• A contradiction in terms: AI interoperability in healthcare Blog post from Jon
• Novel Techniques to Assess Predictive Systems and Reduce Their Alarm Burden Paper by Jon and others
• Shocking new discovery: Recall and Sensitivity are not the same! Blog post summarizing Jon's paper 
• Can a “goldilocks” Mortality Predictor Perform Consistently Across Time and Equitably Across Populations? Pre-print paper from Jon and othersl
• Dwight
• The case for information fiduciaries: The implementation of a data ethics checklist at Seattle Children’s Hospital Dwight et al

Click here for the full archive of town hall recordings. The archive is available only to HDAA members.

Want to learn more?

For more information, or if your organization is interested in presenting, contact Bruce Douglas HDAA Programs