AdvaMed makes recs on using AI in health care 

WASHINGTON – AdvaMed has responded to a request for information from U.S. Rep. Ami Bera, D-Calif., on the use of artificial intelligence in health care. In a letter to Bera, the group outlines measures that could ensure AI is used safely and reliably in health care, including: 

• Improving policies for access to data sets necessary to develop effective AI algorithms and promote AI innovation.   

• Expanding FDA’s authorization to permit use of electronic labeling (e-labeling) in lieu of paper labeling. e-labeling, particularly for software-based devices, is a more efficient and environmentally conscientious method to transfer information to the user. Further, e-labeling offers flexibilities in the presentation of information that can promote health care equity. For example, e-labeling font size can be adjusted for the visually impaired and can be provided in multiple languages and formats.  

• Revising the current FDA guidance on Clinical Decision Support Software to ensure the recommendations better align with the 21st Century Cures Act (e.g., ensure AI-based CDS that produces a single output, such as a recommendation for a particular treatment option that is consistent with common treatment guidelines, wouldn’t de facto be regulated as a medical device). 

“We recognize that the pace of innovation is fast, but we believe the FDA’s existing regulatory authorities are robust and flexible to ensure the safe and effective use of AI/ML technology in medical devices,” AdvaMed stated in the letter. “As state and federal legislators seek to ensure AI/ML-enabled technologies in all industries are used safely, FDA should continue to maintain regulatory oversight of medical devices, including the interoperability between devices and non-devices.” To read AdvaMed’s letter in full, go here.