Marketing Submission Recommendations for a Predetermined Change Control Plan (PCCP) for Artificial Intelligence/Machine Learning (AI/ML)-Enabled Device Software Functions
Are you curious about the latest developments in regulations in the field of AI/ML-enabled device software functions? If so, you'll be interested to know that the Food and Drug Administration (FDA) has released a draft guidance document on the marketing submission recommendations for a predetermined change control plan for AI/ML-enabled device software functions. This guidance document is open for public comment here until July 3rd, 2023, and the FDA is actively seeking feedback from industry professionals.
As AI/ML-enabled devices become more prevalent in the healthcare industry, it's important to ensure that these products meet the highest safety and effectiveness standards. The FDA's draft guidance document provides recommendations on how to develop a predetermined change control plan for AI/ML-enabled device software functions. This will help manufacturers better understand how to manage software changes to ensure that the device continues to function safely and effectively.
By understanding the recommendations and providing feedback, you can help ensure that these devices are safe, effective, and meet the needs of patients and healthcare providers. So let's dive into the FDA's draft guidance and see what it means for the industry!
The key components for PCCP that the draft guidance suggests are :
- Description of Modifications
- Modification Protocol
- Impact Assessment
1. Description of Modifications: This section should specify the planned modifications to the device, including changes to the machine learning model, that are authorized by the PCCP. The Description of Modifications section of a Predetermined Change Control Plan (PCCP) for medical devices that use machine learning models should include a detailed description of each planned modification to the ML-DSF that the manufacturer intends to implement.
The detailed description should describe changes to the device characteristics and performance resulting from the implementation of the modifications. The Description of Modifications should enumerate the list of individual proposed device modifications discussed in the PCCP, as well as the specific rationale for the change to each part of the ML-DSF that is planned to be modified. In some situations, a Description of Modifications will consist of multiple modifications. It may be helpful to reference the labeling changes that are associated with each modification in the Description of Modifications section.
2. Modification Protocol: This section should describe the methods that will be followed when developing, validating, and implementing the modifications specified in the Description of Modifications section. It should include verification and validation activities, including pre-defined acceptance criteria, to ensure that the modifications are safe and effective
Medical devices that use machine learning models should include four components: data management practices, re-training practices, performance evaluation protocols, and update procedures for each modification in the Description of Modifications for the machine learning device. The data management component should include collection protocols for new training and testing data, while the re-training component should describe how the machine learning model will be re-trained with new data. The performance evaluation component should describe how the performance of the machine learning model will be evaluated after modifications, and the update procedures should describe how the modifications will be implemented, how communication regarding the device updates will be provided to users, and how potential risks associated with the update process will be mitigated. The Modification Protocol should also be traceable and specific to the modifications detailed in the Description of Modifications section and sufficiently comprehensive to support specific modifications.
1. Data Management Practices: This component should describe how the manufacturer will collect and manage data for training and testing the machine learning model updates. It should include information on the inclusion/exclusion criteria for data collection, the intended distribution of the data set along covariates describing the patient population and data acquisition conditions, whether the data will be collected prospectively or retrospectively, and any plans for enrichment or stratified sampling to include specific patient subgroups
2. Re-Training Practices: This component should describe how the manufacturer will retrain the machine learning model after making modifications to it. It should include information on the retraining data set, the retraining algorithm, the retraining frequency, and the retraining acceptance criteria
3. Performance Evaluation Protocols: This component should describe how the manufacturer will evaluate the performance of the machine learning model after making modifications to it. It should include information on the performance metrics, the performance evaluation data set, the performance evaluation algorithm, and the performance evaluation acceptance criteria
4. Update Procedures: This component should describe how the manufacturer will update the machine learning model after making modifications to it. It should include
3. Impact Assessment: This section should document the assessment of the benefits and risks of implementing the proposed PCCP, as well as the plan for risk mitigation
The manufacturer's existing quality system should be used as the framework in which to conduct an Impact Assessment for the modifications set forth in the PCCP. The Impact Assessment documentation should discuss how the individual modifications included in the PCCP impact not only the machine learning-based medical device but also how they impact the overall functionality of the device, including how they impact other device software functions, as well as device hardware, if applicable. Additionally, the Impact Assessment should discuss how the implementation of one modification impacts the implementation of another, and the collective impact of implementing all modifications. The Impact Assessment helps to tie the Description of Modifications to the Modification Protocol in that the Impact Assessment identifies the benefits and risks introduced by the specified, planned modifications and how the verification and validation activities of the Modification Protocol will continue to assure the safety and effectiveness of the device.
Use a QMS feature rich for SaMD and AI-ML Teams
Sierra QMS is designed for organizations that are looking to market solutions for the Healthcare and Life Science spaces and is built with features rich for SaMD and AI teams. It is built for engineering teams to operate with their preferred toolsets while automating compliance with QMS principles for global markets.
Along with the system's customization capabilities, our Compliance Experts are accessible for support throughout your entire implementation process to ensure smooth adoption and future usage of your QMS. Sierra Labs is the perfect partner for any startup looking to implement an efficient eQMS that complies with global regulations. We can establish procedures for regulatory compliance tailored to the specific markets you’re interested in entering.
It's that simple. No CC required.