The recently published guiding principles for the development of predetermined change control plans (PCCPs) for machine learning-enabled medical devices (MLMDs) by the MHRA, FDA, and Health Canada aim to support manufacturers by reducing the regulatory burden of reassessment following changes and updates to their devices.
In the UK, when manufacturers make significant updates or changes to their medical devices, they must notify their conformity assessment body and undergo reassessment to ensure performance and safety have not been negatively impacted. However, for medical devices using AI and machine learning, frequent updates may result in lengthy reassessment processes, creating a regulatory burden for developers and assessors.
To address this challenge, the introduction of PCCPs allows manufacturers of MLMDs to demonstrate the specific changes and updates they intend to make, along with their plans to ensure safety and effectiveness are maintained. This eliminates the need for regulatory intervention and streamlines the assessment process.
The five guiding principles for MLMD manufacturers emphasize the need for focused and bounded plans, driven by a risk-based approach that adheres to the principles of risk management. Manufacturers must also provide evidence that the benefits of the changes outweigh the risks throughout the product lifecycle. Transparency is crucial, with clear and appropriate information shared with stakeholders, including patients and healthcare professionals. Additionally, manufacturers should consider the perspectives of all stakeholders to improve the quality and integrity of their PCCPs.
These guiding principles align the expectations of the MHRA, FDA, and Health Canada, reducing or removing the need for reassessment. While they support PCCP development across the UK, US, and Canada, each regulator will have specific national guidance that manufacturers must follow.
Dr. Paul Campbell, the MHRA head of software and AI, emphasizes the importance of adapting regulatory processes to support innovations in AI and MLMDs while ensuring patient safety. The collaboration between regulators on these guiding principles showcases how working with international partners can lead to agile regulatory processes that benefit manufacturers and patients globally.
Overall, the introduction of PCCPs for MLMDs is a significant step in reducing the regulatory burden on manufacturers and streamlining the assessment process for AI and machine learning-enabled medical devices. By providing clear guidelines and aligning expectations, regulators can support innovation while maintaining safety standards.