New FDA Draft Guidance on Predetermined Change Control Plans for AI/ML-Enabled Devices

What You Need to Know for FDA Clearance

Guidance on Predetermined Change Control Plans for AI/ML-Enabled Devices

FDA recently released new draft guidance regarding Predetermined Change Control Plans (PCCPs) for Artificial intelligence / Machine learning (AI/ML) – enabled devices software functions.  If you have a device that has AI/ML functionality in it, or are planning on developing one, and want to submit that device for FDA clearance/approval, you should review this draft.  The draft guidance applies to devices that would normally require a premarket approval or supplement.

Why has this document been released for review?

First, FDA will remind you that they have a commitment to ensuring the safety and effectiveness of health technologies.  This includes “emerging” technologies like artificial intelligence (AI) and machine learning (ML).  They realize that AI/ML has become an important part of medical devices and shows no signs of going away.  They also understand, that by their nature, the development of these products is an iterative process.

Secondly, FDA has the authority to do this. As part of the Food and Drug Omnibus Reform Act of 2022, FDA was given the authority to approve or clear Predetermined Change Control Plans (PCCPs). This draft guidance is one of the first applications of this authority and covers AI/ML-enabled devices requiring premarket approval or premarket notification. Further, FDA may require that those PCCPs include labeling for safe and effective use of the device.

It’s important to note here that while guidance documents are not legally-enforceable, they do give you a look “into the mind” of FDA and lets you know what they’re thinking and what attitudes they currently have on the subject.  In our experience, there aren’t many significant changes between a draft guidance and a formally-released guidance.

The draft guidance provides recommendations on the information to be included in a Predetermined Change Control Plan (PCCP) provided in a marketing submission for devices that includes AI/ML-enabled device software function.

The draft guidance recommends that a PCCP describe the planned device modifications; the associated methodology to develop, implement, and validate those modifications; and an assessment of the impact of those modifications.  The PCCP is reviewed as part of a marketing submission to ensure the continued safety and effectiveness of the device without necessitating additional marketing submissions for implementing each modification described in the PCCP.

The benefit of submitting a PCCP is that, as long as you stick to the PCCP, there’s no need to make additional submissions for each future modification.  Otherwise, changes that could affect the safety or effectiveness of a device would require premarket authorization.

510(k) submission

An important note for anyone interested in submitting a 510(k) and would want to use a device that was authorized with a PCCP as your primary device.  As part of your submission where you make the argument for substantial equivalence between your device and the primary device, you can compare to the version of the device that was cleared or approved PRIOR to changes that the primary device has made under their PCCP.

In this draft, FDA provides recommendations in three areas:

1) detailed descriptions of the specific, planned device modifications;
2) methodology to develop, validate, and implement those modifications in a manner that ensures the continued safety and effectiveness of the device across relevant patient populations (they’re referring to this as a “Modification Protocol”);
3) Impact Assessments to describe the assessment of the benefits and risks of the planned modifications and risk mitigations.

 

There’s a lot to “unpack” in this draft, but it’s important to see where FDA is going and how this can affect how you develop and submit your device(s) going forward.

So, things you should do:

  • Download the draft guidance and read it and understand key points. 
  • If you have strong opinions about items that should be changed, let FDA know about it.
  • Take elements of guidance into consideration as part of your device development.
  • If you’re submitting to FDA for a product with AI/ML functionality, you can engage FDA with pre-submission (Q-Sub) meeting for free prior to submission.
  • If you have questions, please reach out to professionals, like the ones we have at CENIT, to get answers.

This article serves as information only and does not represent an official or agreed position of CENIT Consulting. The opinions expressed in the article are solely those of the authors. Although every effort has been made to ensure the accuracy of the content, CENIT Consulting cannot be held responsible for any loss or damage that may result from relying on the information provided, except where such liability cannot be excluded by law.