The Food and Drug Administration (FDA) has issued guidance on what to include in a predetermined change control plan (PCCP) for marketing AI-enabled medical devices. Highlighting its commitment to innovative regulatory approaches, the FDA stated the guidance supports iterative improvements to AI-enabled device software functions (AI-DSFs) while ensuring safety and effectiveness. It offers recommendations on suitable modifications for a PCCP but does not provide an exhaustive list.
Manufacturers are encouraged to use the Q-Submission Program for FDA feedback on proposed PCCPs before marketing submissions, though PCCPs are not approved during pre-submission. Once authorized, PCCPs require manufacturers to update device labeling to inform users about software updates and their impact on performance. For devices with machine learning, labeling must note that updates could modify functionality.
A PCCP should clearly describe planned modifications, specifying performance changes that meet protocol requirements without requiring new submissions. The FDA recommends detailing individual modifications and their rationale, noting that some PCCPs may involve multiple changes.
While guidance documents are not legally binding, they reflect the FDA’s current thinking and apply to AI-enabled devices reviewed under the 510(k), De Novo, and PMA pathways. The guidance aims to foster the development of safe, effective AI-enabled devices as AI and machine learning advance in healthcare.
In October, the FDA outlined its vision for AI regulation, emphasizing alignment with international standards due to the global reach of U.S.-regulated products. Similarly, the EU AI Act, effective in August, sets rules for developing and using AI in Europe, targeting high-risk applications such as medical devices, biometric identification, and healthcare access. The Act promotes trustworthy, human-centric AI while ensuring health, safety, and fundamental rights protections.
Click here to read the original news story.