AdvaMed Issues AI Policy ‘Roadmap’ for Regulators

AdvaMed, the Medtech Association, released its “AI Policy Roadmap,” a detailed policy outline for U.S. lawmakers and regulators to consider in advancing safe, effective artificial intelligence (AI)-enabled medical and digital health technologies. 

“The future of AI applications in medtech is vast and bright. It’s also mostly to be determined. We’re in an era of discovery,” AdvaMed President/CEO Scott Whitaker said. “While none of us can anticipate all the game-changing applications of AI in medtech to come, we can confidently predict that transformation will continue at a rapid pace—and the policy environment absolutely must keep up. This is the right time to promote the development of AI-enabled medtech to its fullest potential to serve all patients, regardless of zip code or circumstance. Our industry looks forward to continuing to work with Congress on these exciting advancements.” 

The policy proposal outlines the background of U.S. Food and Drug Administration (FDA)-authorized AI-enabled medical devices – more than 1,000 over the last 25 years—and includes numerous examples of real-life instances in which AI-enabled health tech is serving patients, including software analyzing digital images to detect prostate cancer, cuffless blood pressure monitoring at home, and insertable cardiac monitors with algorithms to improve diagnoses of abnormal events. 

Developed by AdvaMed’s Digital Health Tech division, the roadmap addresses five key policy areas: privacy and data access, FDA AI regulatory framework, reimbursement and coverage, AI assurance labs, and generative AI.

“Recent breakthroughs in machine learning and AI are already transforming health care—streamlining administrative workflows, minimizing scan times, monitoring patient health both in and out of clinical settings, and significantly reducing wait times. This is just the beginning. We see tremendous potential to harness AI not only to automate routine tasks, but also to deliver deeply personalized care and treatment,” said Dr. Taha Kass-Hout, global chief science and technology officer, GE HealthCare, and chair of AdvaMed’s Digital Health Tech division board. “The continued support of Congress and the Administration will be important in creating policies necessary to help further ensure as many patients as possible have access to AI-enabled innovations to speed diagnosis, promote the right treatment, and increase access to high quality and affordable care.”

The policy recommendations are as follows:

1. Privacy and Data Access. Data quality and provenance are important considerations for AI medical device developers in training and validating AI models. Privacy law requirements for de-identification and/or minimization of personal data or metadata can be inconsistent and at odds with these important considerations. These inconsistencies impact the ability of AI medical device developers to access, store and retain training and validation datasets (and metadata) over a certain period of time to meet FDA’s recommendations; and demonstrate that the dataset used to train and tune the device is robust and representative of the intended patient population.
   • Recommendations:
     • Ensure data protection without stifling innovation.
     • Evaluate the need to update HIPAA for the AI era and create clear guidelines specifically for data use in AI development.
     • Develop appropriate guidelines around patient notice and authorization for the data used to develop AI.
2. FDA AI Regulatory Framework. FDA’s oversight is guided by a risk-based framework that includes a rigorous premarket review process as well as extensive post-market monitoring requirements after devices are authorized for sale. FDA’s regulatory frameworks are able to accommodate emerging technologies in medical devices, such as AI. In 2022, Congress passed predetermined change control plan (PCCP) legislation. The PCCP framework has great applicability to all medical devices. For AI-enabled devices, in particular, it enables the regulatory framework to keep better pace with the rapid innovation inherent to AI, thereby ensuring clinicians and patients have timely access to improved devices.
   • Recommendations:
     • FDA should remain the lead regulator responsible for overseeing the safety and effectiveness of AI-enabled medical devices.
     • The agency should implement the existing PCCP authority to ensure it achieves its intended purpose of ensuring patients have timely access to positive product updates.
     • FDA should issue timely and current AI guidance documents related to AI-enabled devices and to prioritize the development and recognition of voluntary international consensus.
     • FDA should establish a globally harmonized approach to regulatory oversight of AI-enabled devices.
3. Reimbursement and Coverage. There is no “one size fits all” reimbursement policy for every AI technology. Instead, appropriate payment mechanisms vary depending on the kind of technology in question and the clinical setting in which it is used. Regardless, accurately capturing the cost and value of these technologies is critical to ensuring appropriate reimbursement.
   • Recommendations:
     • Consider legislative solutions to address the impact of budget neutrality constraints on the coverage and adoption of AI technologies.
     • CMS should develop a formalized payment pathway for algorithm-based health care services (ABHS) to ensure future innovation and to protect access to this subset of AI technologies for Medicare beneficiaries.
     • To ensure future innovation and to protect access to ABHS for Medicare beneficiaries, CMS should develop a formalized payment pathway for ABHS.
     • Facilitate the adoption and reimbursement of digital therapeutics through legislation and regulation.
     • CMS should leverage its model authority to test the ability of AI technologies to improve patient care and/or lower costs.
4. AI Assurance Labs. The FDA and other health care stakeholders are exploring the creation of third-party quality assurance labs to support AI lifecycle management in medical devices, with a focus on ongoing performance monitoring. However, concerns have arisen about the suitability and practicality of this approach. Issues include data sensitivity as well as the unclear and evolving scope of these third-party labs’ roles.
   • Recommendation:
     • Policymakers should encourage FDA to participate in the development of and timely recognition of accredited and consensus-based standards for quality assurance processes rather than rely on third-party assurance labs.
5. Generative AI. Emerging use cases for generative AI in medical imaging and other medical applications underscore the need for regulatory flexibility. As emphasized in recent industry comments, any future regulatory approach should carefully weigh the benefit-risk profile of these devices, recognizing the FDA’s existing robust framework for evaluating technology safety and efficacy throughout the product lifecycle. Compared to other types of AI, generative AI’s dynamic outputs may present unique regulatory considerations.
   • Recommendations:
     • Ensure FDA reviews and considers GenAI-enabled devices using the existing risk-based frameworks.
     • Encourage FDA to maintain ongoing dialogue with stakeholders in the health care sector and regular information-sharing on generative AI applications in medical devices.

Congress already has already toyed with the idea of AI-enabled health tech, setting the stage for future policymaking. Last year, the U.S. House of Representatives’ bipartisan Task Force on Artificial Intelligence urged the Centers for Medicare and Medicaid Services (CMS) to develop a formalized payment pathway for AI-enabled medical devices. Kass-Hout contributed to a task force panel. The bipartisan Senate Artificial Intelligence Caucus has hosted sessions on Capitol Hill including an AI demonstration day to show lawmakers and their staff the latest AI-enabled health tech. 

AdvaMed’s recently released Medical Innovation Agenda for the 119th Congress includes recommendations on harnessing the power of AI to improve patient outcomes and access to innovative medtech.