CDRH's 510(k) Working Group—Findings 3 & 4a

A GLIMPSE OF THE FUTURE?
A Cycle of 510(k) Whitepapers by Nancy J Stark and
Kathleen Johnson

The Center for Devices and Regulatory Health has two important missions: 1) to protect the public health and 2) to foster medical device innovation. Their difficult job is to keep these missions in balance, never compromising one in favor of the other.

In August 2010, CDRH's 510(k) Working Group published a preliminary report consisting of more than 60 recommendations grouped under seven findings aimed at improving the Center's effectiveness in implementing its missions. Will the recommendations work? Do they comprise a forward-thinking strategic plan for the device industry? Herein you'll find CDG's summation of Finding 3 and part of Finding 4, along with our opinions of their worth. If you wish to comment, please do so at our blog.

 
Finding 3: the de novo pathtway is important and has not been optimally used across the Center.

The de novo process is an essential tool but difficult-to-use tool
The de novo classification process was established under FDAMA to allow for the 510(k) clearance of devices for which there is no predicate but whose risks do not warrant a PMA approach. The problem with the current system is that a device must first be classified as a Class III through a formal process. The issue is how to expedite classification.

The Working Group recommendation
CDRH reform its implementation of the de novo classification process to provide a practical, risk-based option that affords an appropriate level of review and regulatory control. While the number of de novo requests increases every year (there were 15 in 2009) the review process has become an incredible 2-3 years!

CDG Comments: We agree with the Working Group recommendation and further recommend that the Agency develop a separate de novo application which would include: 1) evidence of no predicate for the intended use and/or technology, 2) demonstration of low risk through design control and biological safety information, and 3) a mechanism for de novo meetings with the Agency. The application should be reviewed within 60 days of submission.

Typically CDRH develops device-specific guidance documents as special controls for de novo devices. CDRH should develop a single generic set of controls that could serve as the baseline. Device-specific guidelines can be added later, as needed.

This would give the Agency an avenue to either quickly classify a device as Class I or II via the de novo process or determine that the device is actually a Class III device.

 

Finding 4: It is challenging for reviewers to obtain the information it needs to make well-supported clearance decisions.

Class IIb Devices versus a "Big Mac" 510(k)
Working Group recommendations
The Working Group recommends creation of a new "class IIb" device classification.

CDG is baffled that the Working Group thinks a new "class IIb" classification is necessary or would even helpful to their obtaining sufficient levels of information about a device to make fully informed decisions.

First, it would add to global confusion. The European Union has a Class IIa and Class IIb device classification system and the EU Class IIb and US Class IIb are unnlikely to be the same. Because most companies do business globally, we should at least be creative enough to think of a new phrase.

And why not build on what we already have: a Special 510(k), Abbreviated 510(k), and Traditional 510(k). If we follow our existing pattern and instead create a new type of 510(k) called a "Super 510(k)" or "510(k) Plus." The Super 510(k) would describe additional information, over and above a traditional 510(k), that would be required for certain, significant-risk Class II devices. A new category of 510(k) can be structured to meet FDA needs, provide more options for the manufacturer, and avoid confusion with the world's second largest device market. We would like to see clear identification of the devices that require clearance under the "Super 510(k)" and defined requirements for data.

Device modifications: to report or not report?
Under current regulations, manufacturers are only required to submit new 510(k)s for device modifications when the modification "could significantly affect the safety or effectiveness of the device" or there has been a "major change or modification to the intended use of the device." The Agency has observed that minor, incremental modifications sometimes accrue to the point that safety, effectiveness, or intended use are affected. Also, there is concern that manufacturers misinterpret the regulation to mean that a new 510(k) is only necessary if a modification definitely does affect safety and/or effectiveness. There is also concern that manufacturers misinterpret the regulation to mean that a new 510(k) is only necessary if a modification could negatively affect safety and/or effectiveness.

When a 510(k) is submitted for a particular modification, CDRH must evaluate not only the effect of that modification itself on the device’s safety and effectiveness, but also the cumulative effect of any unreported modifications that preceded it. There is no explicit requirement for submitters to list the modifications that have been made since the last clearance. This can create challenges for review staff and may result in requests for additional information in the midst of a review.

Working Group recommendations
The Working Group recommends CDRH revise existing guidance to clarify what types of modifications do or do not warrant submission of a new 510(k), and, for those modifications that do warrant a new 510(k), what modifications are eligible for a Special 510(k) (i.e. declaration of conformity to design controls in lieu of providing detailed performance data to FDA as part of a 510(k).) The guidance should clarify that a new 510(k) is also required for changes that potentially could have a positive effect, because what was expected to have a positive impact may in fact create new risks that would not be detected without an adequate assessment.

Finally, the Working Group recommends that CDRH require each manufacturer to provide regular, periodic updates to the Center listing modifications made without 510(k) submission.

CDG comments: Modernizing guidances as to what type of modification requires a new 510(k) submission would be useful. When a 510(k) is submitted, it is reasonable to expect the submitter to identify all interim modifications in both design and/or labeling made since the previous submission.

Periodic modification reporting would result in a huge work overload for both manufacturers and reviewers. Neither party has the human or financial resources to manage regular reporting.

CDG can write your 510(k)

CDG's extensive Network Staff has expertise in writing 510(k)s in a wide variety of therapeutic, management, or diagnostic medical device areas. My co-author, Kathleen Johnson has more than 12 years experience in medical device regulatory affairs. CDG has capabilities in regulatory submissions, clinical research, toxicology, biostatistics, information research, medical writing, design control—we focus exclusively on medical device pre-approval issues.

Our style is to work collaboratively with a point-person on your side so that you are involved in the process every step of the way. Phone or email us at 773-489-5721 or [email protected].

Quality of submissions—crafting an argument
In this section the Working Group takes great pains to avoid stating outright that some submitters can't write: we do not make logical and scientifically developed arguments for safety and effectiveness (i.e. we don't apply the scientific method when structuring our arguments), we are disorganized, and we omit information that is clearly required.

CDRH reviewers state: "...information contained in the “Device Description” section of a 510(k) is often a general summary, and it frequently does not provide an adequate level of detail about the new device...." and "...or many 510(k)s, information provided in one section is inconsistent with information provided in another."

Working Group recommendations
The Working Group recommends the agency adopt the use of an "assurance case" framework for 510(k) submissions. An assurance case is defined as a documented body of evidence that provides a convincing and valid argument that a specified set of critical claims regarding a device's safety and performance are adequately justified. The definition parallels three steps of the scientific method.

 
Further, the Working Group recommends that both photographs and schematics be included with a submission, and that manufacturers retain at least one unit of a device and make it available to the agency for current or future reviews (e.g. when the device is cited as a predicate.)

CDG comments: FDA should update the 510(k) Manual (FDA publication 95-4158) and make it available under "guidances" in Device Advice. This document provides detailed guidance of what each section of a 510(k) should include. There is no need to develop yet another model, when a sufficient and familiar model already exists. Kathleen has used this manual as her guide in developing numerous, successfully cleared 510(k)s.

 
We searched for "510(k) Manual" in Device Advice and it isn't there. We searched for "510(k) Manual" in Google and it isn't there. Kathleen had a copy in her library so we could find the FDA publication number. When we searched for "FDA publication 95-4158" in Google we found a copy under FDA dockets. We were amazed that the 510(k) Manual was so difficult to find and not available at all under Device Advice.

FDA has a website for most of its databases, but the database for guidance documents—surely a key database for good 510(k) submissions—is not there. The website is called FDA Databases and you can see from the image above that there is no tab for guidances!

Part of the problem is that new companies can't find the information they need to tell them how to prepare a 510(k). Maybe FDA should sponsor a public workshop, live or available free as webcasts, to educate new submitters about requirements and expectations for 510(k) submissions.

Listen to Experts Janice Hogan and Yarmela Pavlovic
Discuss "The Successful 510(k)"

As useful as when it was first recorded, Janice Hogan and Yarmela Pavlovic present the whys and hows of crafting a successful 510(k). Listen on CD or OnDemand and take advantage of their vast experience. Their detailed and thorough presentation covers: 1) the legal framework, 2) finding predicates, 3) choosing the right claim, introducing new features, 4) assessing data requirements, 5) software and combination products, 6) dealing with a not-substantially equivalent (NSE) determination, 7) handling device modifications, and 8) seeing the 510(k) from the reviewers perspective. You'll receive the slide deck, video/audio recording, and links to important guidances. Order here.

CDG disagrees that exemplar devices be retained for current or future reviews. Some devices cost many thousands of dollars, some fill an entire room, and some are not stable. Retaining examples would be unduly burdensome and in some cases an impossibility. Furthermore, one manufacturer should not be compelled to make an example device available to FDA so that FDA can have it available for reviewing a submission from another manufacturer.

Improper use of recognized standards—require a summary of testing
Section 514(c)(1) of the Act describes the possibility of using, as part of a 510(k) submission, consensus standards that have been entirely or partially recognized by FDA. This section states, “[FDA] shall, by publication in the Federal Register, recognize all or part of an appropriate standard established by a nationally or internationally recognized standard development organization for which a person may submit a declaration of conformity in order to meet a premarket submission requirement or other requirement under this Act to which such standard is applicable.”

Working Group Issues
The problems are that: 1) a device may raise questions of safety and performance not addressed by the consensus standard; the standard provides only partial evidence of substantial equivalence, 2) submitters may not use the most current version of a standard, they may not be able to determine which sections of the standards are recognized and which are not (you must turn to the supplemental information sheets), and 3) submitters do not always indicate how they have used the standard or deviated from the standard.

CDG Comments: The recognized version of a standard, not the current version, should be used. The 510(k) Manual should be updated to discuss the proper use of consensus standards and the "Standards Data Report Form for 510(k)s" should be modified to provide a place for summary information and all other relevant information FDA requires.

Where's the rest? The last five recommendations under Finding 4 will come out next week. Got something to add? Please comment below.

Best Regards,
Nancy J Stark, PhD
President, Clinical Device Group Inc 

 

Comments

[Nancy J Stark]

Dear Mr. Doyle,
Thank you for your comment. To be honest, I spent many hours looking for the origin of CDRH’s “mission” to foster medical device innovation. I found hundreds of references to it, including the one I finally used from CDRH’s website; but I could not find the original citation.

I went back to an original 1976 version of the Food, Drug and Cosmetic Act that I had laying around, I checked the current version of the Act and found the paragraph you quote in your comment, but I never did find an original citation for “foster medical device innovation”.

Your correction is welcome. And I wonder, do you have an idea of why so many people—FDA and the Working Group preliminary report people included—believe that fostering innovation is part of CDRH’s job description? How did this idea get started?