On April 3, 2023, U.S. Food and Drug Administration (FDA) issued its much anticipated draft guidance, "Marketing Submission Recommendations for a Predetermined Change Control Plan for Artificial Intelligence/Machine Learning (AI/ML)-Enabled Device Software Functions" (Draft Guidance).
While FDA has already approved, authorized, or cleared over 500 AI/ML devices, FDA continues to receive an increasing number of marketing submissions and pre-submissions for AI/ML-enabled medical devices. FDA’s traditional approach for the regulation of hardware-based medical devices, however, is not well suited for the faster, iterative design and development, and type of validation used for software device functions, including Software as a Medical Device.
The benefit of AI/ML is that it can optimize performance over time and continuously learn based on real world experience. Under the traditional FDA regulatory framework, changes to software require new risk assessments to determine whether the change affects the functionality or risk category before releasing each change. That is, the algorithm is essentially locked and cannot change while out in the market, defeating the optimization of AI/ML technology.
To address this issue, FDA outlined a Predetermined Change Control Plan for premarket submissions in its 2019 Discussion Paper and 2021 AI/ML-based Software as a Medical Device Action Plan, allowing manufacturers to predict algorithm changes and implement future modifications without requiring additional marketing submissions.
Under a Predetermined Change Control Plan, manufacturers would be required to submit:
The Draft Guidance builds on the proposed framework and helps clarify the types of modifications that should be included in the Predetermined Change Control Plan. Notably, FDA also proposes that the Predetermined Change Control Plan articulated in the initial proposed framework be used for not only AI/ML-enabled Software as a Medical Device, but for all AI/ML-enabled device software functions, including software functions that are part of or control hardware medical devices.
Under this framework, FDA expects manufacturers to commit to transparency and real-world performance monitoring, and to periodically update FDA on changes implemented as part of the approved pre-specifications and algorithm change protocol. In addition, modifications should be implemented following appropriate, well-defined practices, such as the Good Machine Learning Practice guiding principles jointly developed by FDA, Health Canada, and the United Kingdom’s Medicines and Healthcare products Regulatory Agency.
Ultimately, the new Draft Guidance seeks to enable manufacturers to market medical devices with continuously learning algorithms without having to obtain a new authorization or clearance for each change, so long as the changes are in line with the predetermined plan. Through this Draft Guidance, FDA seeks to provide flexibility for devices with continuously learning algorithms, while retaining certain limits on the software to safeguard continued safety and effectiveness of the devices.
FDA invites feedback on this Draft Guidance. Comments can be submitted until July 3, 2023.
Foley is here to help you address the short- and long-term impacts in the wake of regulatory changes. We have the resources to help you navigate these and other important legal considerations related to business operations and industry-specific issues. Please reach out to the authors, your Foley relationship partner, or to our Health Care Practice Group with any questions.