MHRA, FDA, and Health Canada Publish Guiding Principles for MLMD Manufacturers
The MHRA (Medicines and Healthcare products Regulatory Agency), FDA (Food and Drug Administration), and Health Canada have jointly released a set of guiding principles for manufacturers of Machine Learning and Machine Learning-based Medical Devices (MLMDs). These principles aim to assist manufacturers in reducing regulatory requirements when implementing changes and updates to their devices.
The guiding principles for MLMD manufacturers build upon the previously established 10 guiding principles for Good Machine Learning Practice. The new set of principles outlines that a Pre-specified Change Control Plan (PCCP) should be focused and bounded, describing specific changes intended by the manufacturer. It should also be risk-based, driven by a risk management approach. Furthermore, the plan should be evidence-based, demonstrating that the benefits of the changes outweigh the risks throughout the product lifecycle. The PCCP should be transparent, providing clear and appropriate information to all stakeholders, and it should adopt a total product lifecycle perspective by considering the perspectives of all involved parties.
Currently, in the UK, when a manufacturer makes significant updates or changes to their medical device, they must inform their conformity assessment body, and their device may be reassessed to ensure that the changes have not compromised its performance and safety. However, medical devices utilizing AI and machine learning may require frequent updates, potentially leading to lengthy reassessment processes with each change. This poses a significant regulatory burden for both developers and assessors.
The introduction of PCCPs allows MLMD manufacturers to demonstrate the changes and updates they intend to make, along with the measures they will take to ensure ongoing safety and effectiveness, eliminating the need for regulatory intervention.
While these guiding principles will apply across the UK, US, and Canada, each regulatory body will provide 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, emphasized the necessity for regulators to adapt their processes to support innovations while ensuring patient safety. By collaborating with the FDA and Health Canada on these guiding principles, the MHRA aims to reduce the regulatory burden on manufacturers and foster agile regulatory processes that benefit both manufacturers and patients globally.
In conclusion, the MHRA, FDA, and Health Canada have released guiding principles to support MLMD manufacturers in reducing regulatory requirements when implementing changes and updates to their devices. These principles aim to strike a balance between supporting innovation and ensuring patient safety. Manufacturers will be able to develop Pre-specified Change Control Plans that outline their intended changes and demonstrate ongoing safety and effectiveness, eliminating the need for reassessment with each update. Through international collaboration, regulators are working towards agile regulatory processes that benefit manufacturers and patients worldwide.
