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 below.
Finding 4: It is challenging for reviewers to obtain the information it needs to make well-supported clearance decisions
[5] 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.
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 areas. My co-author, Kathleen Johnson is a regulatory consultant with more than 10 years experience in medical devices. 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 cdginc@clinicaldevice.com.
[6] Incomplete information—how much is enough?
Current regulations require 510(k) submissions to provide "data to support the statement" regarding substantial equivalence to a predicate device. Under PMA regulations, PMA applications are explicitly required to contain a summary of all information known concerning the safety and/or effectivenes of the device under review. FDA's concern is that submissions may contain biased information, including only clinical or scientific information that supports device equivalency.
Working Group recommendations
The Working Group recommends that 21 CFR 807.87 be revised to explicitly require 510(k) submitters to provide a list and brief description of all scientific information regarding the safety and/or effectiveness of a new device known to or that should be reasonably known to the submitter.
CDG comments: The recommendation to review all available literature is not in keeping with the original intention of the Act, which was to demonstrate substantial equivalence to a predicate. It sounds like a Clinical Evaluation Report for every medical device in Europe. The result could inundate FDA with so much information that an already stressed Agency would increase its review times rather than become more efficient.
Summarizing data generated by other parties on similar, but not equivalent devices, is not helpful; summarizing data generated by other parties on equivalent devices but dissimilar test methods is not helpful. Much information is developed using early design or prototype devices. Would the recommendation require a review of every document in the design history file and before? It has the feel of a PMA, where safety and performance are established de novo.
Listen to Experts Janice Hogan and Yarmela Pavlovic
Discuss "The Successful 510(k)"Listen to experts Janice Hogan, Esq and Yarmela Pavlovic, Esq. present the whys and hows of crafting a successful 510(k) 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) chosing 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.
[7] Type and level of evidence needed
The reviewer's first step is to determine if the device and predicate have the same intended use. However, once this is done, most devices have either different indications for use and/or different technological characteristics than the predicate, and FDA may want data to show that any differences have not altered the intended use or significantly affected the safety and/or effectiveness of the device. In most cases data from bench and/or animal testing is sufficient to complete the argument of safety and effectiveness, but in some cases the Agency wants additional data.
Working Group recommendations
The Working Group recommends the agency develop guidance defining a subset of devices, i.e. Class IIb devices, for which clinical information, manufacturing information, or post-marketing information would be typically necessary.
CDG has already stated that it disagrees with defining Class IIa and Class IIb devices because it will lead to confusion with European Class IIa and IIb device classification schemes. If clinical, manufacturing, or post-market information is necessary to successfully argue safety and effectiveness, the agency can address this via a "Super" 510(k). As prevsiouly stated, the regulation numbers of devices and the required information for Super 510(k)s should clearly defined; the information requirements cannot change with each submission.
[8] Clinical information
The Working Group acknowledges there is common misunderstanding as to what constitutes clinical data. Yet within the space of four paragraphs on pages 76-77 of the report the Working group refers to "clinical information", "clinical data", "clinical evidence", "clinical investigation", and "clinical study". According to CDRH staff, submitters sometimes cite the definition of “valid scientific evidence” together with the “least burdensome” provisions of the Act, claiming that it is sufficient for them to provide any type or level of evidence that meets the definition of “valid scientific evidence” with the lowest possible level of burden, even if such evidence fails to provide "reasonable assurance" of safety and effectiveness.
Working Group recommendations
The Working Group recommends a “Class IIb” guidance provide greater clarity regarding the circumstances in which the agency will request clinical data in support of a 510(k), what type and level of clinical data are adequate to support clearance, and to define clinical data. Would this mean that all IVD 510(k)s are Class IIb?
CDG recognizes there are situations in which new clinical data may be necessary to establish substantial equivalence and demonstrate safety and effectiveness.
For the most part, however, substantial equivalence, safety, and performance can be clearly established with valid scientific evidence obtained from bench and animal data. Needlessly requesting new human clinical data is unethical. All clinical trials expose human subjects to risk by their very definition. It is not that not conducting a clinical trial should be justified, but that the very act of conducting a clinical trial should be justified.
By needlessly raising the bar to 510(k) clearance, FDA neglects is obligation to foster medical device innovation.
With regard to clinical information, FDA should harmonize its definitions of clinical information with Europe and GHTF so that communication between submitters and reviewers can be clear. In the Europe and GHTF world clinical information is data from the literature, research in humans is a clinical investigation (study or trial may be used colloquially), clinical evaluations are reviews of clinical information and clinical investigations and the risk management system altogether.
[9] Postmarket information
CDRH feels that for certain devices, including novel or particularly complex technologies, it is not feasible to conduct a large-scale clinical trial prior to clearance [does this mean such devices are now cleared without adequate clinical data?] They feel it is necessary to collect additional data after clearance in order to better evaluate the safety and effectiveness of a device over a longer time period, or in a wider patient population.
CDRH has the authority to require postmarket surveillance studies (also called Section 522 studies) as a condition of 510(k) clearance for devices that are expected to have significant use in pediatric populations, but there is no explicit authority for CDRH to order Section 522 studies as a condition of premarket clearance in other situations.
Working Group recommendations
The Working Group recommends the Agency seek greater authority to require postmarket surveillance studies as a condition of clearance for certain devices.
CDG comments: The public health may benefit from postmarket registry studies for certain long-term use devices or devices with very long-lasting effects, however the decision to conduct such studies should be that of the manufacturer and not a condition for clearance.
The agency should continue to develop a unique device identification (UDI) system. There is a vast store of information collected everyday from Medicare billing records, Medicaid billing records, the Health Care Utilization Program (HCUP), and other public information sources, which will allow "real-world" anonymized data on device use and outcomes to be searched, sorted, and correlated.
[10] Manufacturing process information
Working Group recommendations
Because 37% of device recalls are associated with manufacturing problems, the Working Group recommends CDRH develop guidance to provide greater clarity regarding what situations may warrant the submission of manufacturing process information as part of a 510(k), and include a discussion of such information as part of its “class IIb” guidance. The Working Group further recommends the agency seek authority to withhold clearance on the basis of a failure to comply with good manufacturing practice requirements in certain situations.
CDG comments: There are situations, as with novel materials, when manufacturing process and design control information should be part of the 510(k) and that such situations should be addressed in a guidance.
A requirement to have evidence of a good manufacturing practice system in place prior to clearance moves the 510(k) process into the arena of a PMA. Many of the proposals by the Working Group blur the line between the 510(k) and PMA processes. The distinction is clear in the Act and should be honored by the Agency.
Best Regards,
Nancy J Stark, PhD
President, Clinical Device Group Inc
We (our manufacturer) recently gave up on a submission after close to a year of frustration dealing with a reviewer who knew nothing about the class of IVD's (coagulation) he was reviewing
which caused substantial wasted time/energy in answering questions and providing additional documentation which were not needed or from our perspective the intent of the a 510k submission
Posted by: Michael Shiflett | 09 December 2010 at 11:41 AM