AI: The health equity advocate I never expected
by Abner Mason
Abner Mason, SameSky Health founder and CEO, and Josh Siegel, SameSky Health chief technology officer
When I conceived of SameSky Health, I didn’t think twice about artificial intelligence (AI). My only goal was to improve experiences, access, and outcomes for communities that healthcare had long overlooked. But along the way, AI became essential to our health equity mission.
That might sound surprising. After all, healthcare has seen AI solutions nosedive after reinforcing harmful biases. Some technologies worsened the health disparities that already cost untold lives and dollars.
Now, as the industry races to implement AI, we face a question: How can we ensure the technology helps rather than harms health equity?
I recently sat down with Josh Siegel, our chief technology officer, to discuss how SameSky Health developed equitable AI.
Our CultureGuide® platform’s AI capability has enabled us to personalize member engagement on behalf of our clients quickly and cost-effectively, expanding healthcare access and boosting outcomes. But too many other organizations have approached the technology without considering the voices, cultures, and experiences of diverse people, which allowed unintentional bias to seep into their strategies.
My conversation with Josh revealed an important truth: Healthcare organizations don’t need to choose between rapid innovation and health equity.
Here’s how.
Equitable AI Starts with the Right Data Set
We positioned ourselves to develop AI through a health equity lens long before we wrote a single line of code.
Throughout 10 years of text conversations and phone calls with people from underserved communities, our team had engaged with millions of health plan members. They discussed everything from COVID-19 vaccinations to Annual Wellness Visits to diabetes support.
Over time, we logged, interpreted, and cataloged these exchanges, with members’ informed consent. We compiled a proprietary data set detailing how culturally diverse people discuss complex health topics — the type of insight that most organizations never attain. As far as we know, it’s the only data set of its kind.
That information, combined with a health-equity-focused data analysis, laid the foundation.
A Commitment to Continuous Improvement
Using our data set, Josh’s team of technologists and data scientists collaborated with Community Health Guides — the human experts who interact with the people we serve — to engineer an AI solution capable of understanding members’ text messages and generating engaging content.
Real people guide the solution’s evolution with health equity at the forefront, moderating its decision-making and mitigating potential biases. They continuously enhance the AI by reviewing its results and training the algorithms on new data.
Our data set, cultural expertise, and commitment to improvement add up to something greater than the sum of their parts: A solution that augments our teams’ ability to create fluid, natural conversations that inspire people to take control of their health.
How AI Advances Health Equity
The SameSky Health AI solution enables our CultureGuide solution to scale personalized, culturally tailored content for diverse communities across the country. That expands access to information and resources that support better outcomes.
First, our AI solution leverages human- and data-driven insights to create multiple iterations of personalized content. For clients, we create the best opportunity to improve member connections and relationships.
Second, we are able to continuously improve and refine the performance of the Pathways that drive CultureGuide outreach based on previous member interactions, as well as accelerate new content development. For clients, we are able to quickly develop new content and field-test it to expand areas of focus (e.g., new Focus Pathways that address specific gaps in care and other health plan priorities).
The most important metrics focus on the people we serve.
A recent client engagement encouraged 19% of members in a Managed Medicaid plan to complete their Annual Wellness Visit over 12 months. That result was three times better than the client expected.
In another client engagement, we reached out to Medicare Advantage members who hadn’t seen a doctor in 16 months or more. After the yearlong program, 58% of members had completed an Annual Wellness Visit.
As Josh and I reviewed these results, a phrase came to mind: “To whom much is given, much is expected.”
The reason we were able to gather such a thorough data set was because people from underserved communities trusted us. We had a responsibility to repay them. The best way was to leverage AI to reduce the inequities at healthcare’s core.
As AI gains traction in healthcare, every organization must consider the high stakes of innovation: Algorithms change people’s lives. We all share the obligation to ensure they do so for the better.
Abner Mason is the founder and CEO of SameSky Health. Abner has spent decades working to reduce barriers to care faced by underserved people nationally and internationally, from the federal to the local level. He has served on President Bush’s Advisory Council on HIV/AIDS, as Chief Policy Advisor to the Governor of Massachusetts, and as part of the Biden-Harris Campaign Policy Committee. He currently sits on the Boards of Manifest MedEx and the California Black Health Network, is a member of United States of Care’s Founders Council, and the American Medical Association’s External Equity and Innovation Advisory Group. He is also the founder of Health Tech 4 Medicaid.
