MHRA, FDA, & Health Canada Release 5 Guiding Principles to Simplify MLMD Regulations

Predetermined change control plans (PCCP) principles aim to ease regulatory hurdles for machine-learning-enabled medical devices (MLMD) developers, freeing resources to enhance patient-focused product performance.

The UK Medicines and Healthcare Products Regulatory Agency (MHRA), in collaboration with the US FDA and Health Canada, has published five guiding principles to streamline the regulation of machine-learning-enabled medical devices (MLMD) manufacturers.

These principles are designed to assist MLMD manufacturers by reducing the regulatory burden associated with reassessment following specific changes and updates to their devices.

To provide clear and efficient guidelines, these principles build on the ten guiding principles for Good Machine Learning Practice. They specify that a predetermined change control plan (PCCP) for MLMD manufacturers should adhere to the following criteria:

1. Focused and Bounded: The PCCP must describe specific changes the manufacturer intends to implement.

2. Risk-based: It should follow a risk-based approach regarding intent, design, and implementation, adhering to risk management principles.

3. Evidence-based: The plan should demonstrate that the benefits of changes outweigh the associated risks throughout the product's lifecycle.

4. Transparent: The PCCP should provide clear and appropriate information and detailed plans for ongoing transparency to all stakeholders, including patients and healthcare professionals.

5. Total Product Lifecycle Perspective: Manufacturers should continually consider the perspectives of all stakeholders to enhance the quality and integrity of their PCCP.

Under current regulations in the UK, when a manufacturer makes significant updates or changes to their medical device, they must notify their conformity assessment body. The device may then be reassessed to ensure that these modifications have not negatively impacted its performance and safety.

Medical devices utilising AI and machine learning (ML) often require frequent updates, potentially resulting in lengthy reassessments every time a change is made. This can create a substantial regulatory burden for both developers and assessors.

To alleviate this burden, PCCPs offer manufacturers of MLMDs a way to demonstrate the changes and updates they intend to make and how they will maintain safety and effectiveness without regulatory intervention.

These guiding principles lay out the areas in which the MHRA, FDA, and Health Canada share expectations regarding an acceptable PCCP, aiming to reduce or eliminate the need for reassessment.

While these principles will facilitate PCCP development across the UK, US, and Canada, each regulator will have specific national guidance that manufacturers must adhere to. The MHRA plans to publish its guidance in 2024.

Dr Paul Campbell, MHRA head of software and AI, said, "By collaborating with the FDA and Health Canada on these guiding principles, we can clearly outline where we align on our expectations for a successful change control plan and help reduce the regulatory burden for manufacturers. Collaboration between regulators on these guiding principles demonstrates how working with international partners can help the development of agile regulatory processes that support innovative manufacturers and patients globally."

The MHRA plays a critical role in regulating all medicines and medical devices in the UK to ensure their functionality and safety. Their work is grounded in robust, evidence-based judgments to ensure the benefits outweigh any associated risks.