Technology and AI in Healthcare: How to be ANTI-Bias

Proposal: As the healthcare system increasingly relies upon technology and artificial intelligence (AI), it is imperative that leaders understand both the promises and the perils of these tools with respect to existing systems of power, privilege and oppression. Many think of technology as neutral, but in the design, development, and implementation of new technology there must be an actively anti-bias approach in order to combat systemic oppression and racism. Anything less results in technology that is participatory in existing systems, and has the potential to exacerbate disparities in health outcomes and experiences with the healthcare system. When used thoughtfully, AI and other digital technologies offer powerful opportunities to reduce bias. It is important for health educators to include AI as a subject of discussion in their curriculum, to ensure future healthcare professionals understand the importance of digital literacy as a social determinant of health for their patients.

This workshop seeks to give people the tools to discuss bias in technology within the healthcare system and health education spaces, and engage with a critical perspective. Participants will leave equipped to help their health systems recognize biased technologies and to advocate for transformative technologies that increase equity and access.

Expertise of the individual:
ACHIEVE Innovation is a group of women with advanced degrees in education and the health sciences. We work on community, state, national, and international forums to advance health equity in clinical and educational spaces.

Workshop structure:
Introductions (10 minutes): Share your favorite new healthcare technology and why.
- Fill out a questionnaire (or online polling) about how familiar people are with healthcare tech
Introduction presentation on technology in healthcare, key terms and how digital tools such as AI are changing the landscape of healthcare (20 minutes).
- Ask the group, what are some new applications of technology in your daily practice?
Mini case study: Review ChatGPT answers to the most commonly asked medical questions on webMD. Compare this to actual clinician responses. See if people can tell the difference.
- Researchers conducted a study comparing physician and chatbot responses to patient questions on social media. Discuss the potential to use this for good, saving clinician time.
Small group discussion and share (30 minutes): what are some of the positive impacts technology has had on healthcare (either personal experience, or at the systemic level).
- Examples:
- Google
- CVS
Introduce some of the potential harms of AI in healthcare through a case study of the Epic predictive model (15 minutes).
- This model predicted higher rates of no show in Black patients. In clinics using this model, the patients with a higher likelihood of no show were double-booked rather than being offered additional support to attend their appointment (victim blaming).
Training for artificial intelligence happens based on data from a biased system; how can we fight against this?
Digital literacy as a social determinant of health
Discussion around how to recognize bias in technology, and what questions to ask when you implement new technology (45 minutes).
What design processes can help us avoid bias in new healthcare technologies, especially AI-based ones?
How do we develop tests for technologies in order to detect inadvertent or intentional biases?
Who is accountable at your organization? How do you hold them accountable?
Allies in this space (15 minutes): Present groups working on equity in technology and resources they offer such as:
- Global Strategy on Digital Health
- The Digital Medicine Society
- Frameworks for Digital Health Equity
Wrap up (10 minutes):
- What are two things you will do in the next month to ensure that your organization is using new technology to combat bias?
Identify resources and tools for further learning