On October 20, 2026, FDA highlighted resources on Predetermined Change Control Plans (PCCPs), including an overview, a list of authorized devices with PCCPs, and hypothetical examples involving novel medical devices. For manufacturers planning product updates, these resources offer a useful starting point for understanding how future changes can fit into a regulatory strategy. [1]
What is a PCCP?
A PCCP is a plan submitted for FDA review that identifies specific future device modifications and explains how they will be evaluated and implemented.
When FDA authorizes the plan, the manufacturer can implement changes covered by it without a separate marketing submission for each change, provided implementation follows the authorized plan. PCCPs may be included in submissions through the 510(k), De Novo, and Premarket Approval (PMA) pathways.
The scope matters: changes must remain within the device’s intended use. For devices subject to 510(k) requirements, substantial equivalence to the predicate device must also be maintained. [2]
What does the plan contain?
FDA’s final guidance for AI-enabled device software functions describes three central elements:
- Description of Modifications: the specific changes the manufacturer intends to make.
- Modification Protocol: the methods for developing, validating, and implementing those changes.
- Impact Assessment: an evaluation of how the changes affect the device, including their benefits and risks.
The guidance supports iterative improvement while maintaining assurance of safety and effectiveness. It applies to AI-enabled devices reviewed through the 510(k), De Novo, and PMA pathways, including relevant device-led combination products. [3]
For a development team, this makes future updates a subject for early planning. A proposed improvement needs a defined scope, an evaluation method, and a clear basis for deciding whether it is ready for release.
Two examples that make the concept practical
FDA’s hypothetical cases illustrate different ways a PCCP could operate.
Adapting a sepsis alert system to a hospital. A manufacturer could plan specified adjustments using local hospital data, such as model fine-tuning or changes to an alert threshold. In FDA’s example, the adapted model is locked before validation, and the manufacturer reviews and approves the adaptation before deployment.
Updating cardiac arrhythmia detection software. A separate example describes periodic model retraining using newly annotated ECG data. Updates could be deployed automatically after successful validation against predefined criteria. If any acceptance criterion is missed, deployment is blocked and the manufacturer is alerted.
These are illustrations, not authorizations of particular products or complete submission templates. FDA explicitly notes that the examples do not contain all the information expected in a submitted PCCP. [4]
Where the flexibility ends
A PCCP authorizes a defined approach to specified modifications. An update outside its scope needs a separate regulatory assessment to determine whether another submission is necessary.
The plan also needs to be operational. FDA’s joint guiding principles with Health Canada and the UK’s MHRA emphasize specific boundaries, risk management, supporting evidence, transparency, and oversight throughout the product lifecycle. They address monitoring performance and having mechanisms to stop or reverse changes that fail specified performance criteria. [5]
For manufacturers, the practical implication is to connect regulatory planning with the teams responsible for testing, release decisions, user communication, and monitoring after deployment.
What manufacturers should do now
AVRA recommends starting with a focused internal discussion:
- Map anticipated updates. Identify the modifications the team realistically expects to make.
- Define evidence needs early. Consider the data, testing methods, and acceptance criteria needed to evaluate each update.
- Assign implementation responsibilities. Establish who will maintain the plan, review results, control releases, and monitor performance.
- Assess whether a PCCP fits the product strategy. Compare the effort required to develop and maintain the plan with the expected value of the planned updates.
A useful first deliverable is a simple internal table linking each proposed change to its purpose, risks, testing approach, and responsible owner. That discussion can reveal whether the planned changes are sufficiently defined to support a PCCP strategy.
What is new—and which guidance applies?
The October update provides additional resources and examples. It does not, by itself, establish a new regulation.
As of October 24, 2026, FDA’s AI-specific PCCP guidance is final, with an August 2025 version. The broader guidance, Predetermined Change Control Plans for Medical Devices, remains an August 2024 draft marked “not for implementation.” Manufacturers should distinguish these documents when selecting references for their submissions. [3,6]
Planning changes to a medical device for the U.S. market? Contact AVRA Medical Consulting at contact@avramed.us to discuss your regulatory planning needs.
Official FDA references
- CDRH News and Updates — October 20, 2026
- Predetermined Change Control Plans
- Final guidance: PCCPs for AI-enabled device software functions
- Novel Medical Devices with PCCPs: Hypothetical Case Examples
- PCCPs for Machine Learning-Enabled Medical Devices: Guiding Principles
- Draft guidance: Predetermined Change Control Plans for Medical Devices