On February 1, 2019, ACT | The App Association’s Connected Health Initiative (CHI) convened a multistakeholder dialogue on the role of artificial intelligence/augmented intelligence (AI) in healthcare in Washington, DC. The purpose of this dialogue was to address the role of machine learning and AI in improving healthcare, preventing hospitalizations, reducing complications, and improving patient engagement. This dialogue was chaired by CHI’s Morgan Reed and Brian Scarpelli and was attended by approximately 40 representatives from across the digital health stakeholder community including industry, medical professionals, patients, and government.

This dialogue was based on nearly a year of work that occurred in the CHI’s Healthcare AI Task Force, which began as a series of open roundtable dialogues to examine key issues within the health AI space. The purpose of these meetings was to advance understanding across the connected health community and to seek cross-community consensus on the path forward for advancing the uptake of AI-driven cutting-edge innovations across care settings, building on the good work of pre-existing efforts by groups such as the Partnership for AI, Xavier University, and the American Medical Association, among others. These community consultations formed the foundation for a set of deliverables for policymakers. CHI then convened this February 1 dialogue to solicit feedback on these documents from the broader digital health stakeholder community.

During the dialogue on February 1, we presented the CHI AI Task Force’s deliverables: (1) a position piece supporting AI’s role in healthcare, (2) policy principles addressing how policy frameworks should address the role of AI in healthcare, and (3) a terminology document targeted at policymakers. These documents are “living” and will be updated at a future date to reflect continued feedback from the community.

During the February 1 dialogue, much of the presentation and discussion centered around the policy principles. Many stakeholders provided valuable input on this document, recommending updates regarding the role of voice in shaping the improved AI-powered user interfaces, how to approach locked versus unlocked algorithms (e.g., in frameworks’ approach to auditing), and the important role of access for patients in successful health AI frameworks.

Moving forward, the CHI plans to continue to advance the public dialogue on the role of AI in healthcare:

  • CHI will make its deliverables publicly available to contribute to the public debate regarding AI in health.
  • CHI will convene further targeted working groups to develop additional recommendations under certain areas raised in the health AI policy principles (e.g., ethics, auditing of algorithms, etc.).
  • CHI will convene further dialogues for stakeholders that may feature guest speakers from the industry, academia, and the U.S. government.
  • Informed by input collected during and after this dialogue, CHI will commence a policymaker education campaign through public briefings and events, meetings, and other means.